Full text of "A grammar of New Testament Greek"

See other formats

—F 5 


ie Sy 


Ses 


- “ων: ἂς 
en a οωνς 


il 


all 


NOWO1 


A GRAMMAR OF 
NEW TESTAMENT GREEK 


PRINTED IN GREAT BRITAIN BY 
MORRISON AND GIBB LIMITED 


FOR 


T ἃ T. CLARK, EDINBURGH 


LONDON } SIMPKIN, MARSHALL, HAMILTON, KENT, AND CO. LIMITED 
NEW YORK: CHARLES SCRIBNER’S SONS 


Part 1. issued in 1919 
Part ii. ae ep 900 
Part iii. yy gy Zo 


PREFACE TO VOLUME IL. 


maw oe 


At last, with the publication of Part iii., the second volume of 
Moulton’s Grammar of New Testament Greek is brought to a 
close. The reader may be reminded that before sailing for 
India in October 1915 Dr. Moulton had finished the MS of 
Parts i. and i1., and had already written the important chapter 
upon Word-Composition for Part 11. His intention was to 
complete Part iii. with a chapter on Word-Formation by Suffixes, 
and to enrich the volume with an introductory chapter on 
New Testament Greek, which would lead up to an Appendix 
on Semitisms in the Greek Testament. In writing this Appendix 
he counted on the collaboration of his colleague the Rev. C. L. 
Bedale, a Semitic scholar of real distinction and great promise. 
Dr. Moulton died in the Mediterranean in April 1917, a victim 
of the ruthless submarine campaign. Mr. Bedale died in a 
military hospital at Cambridge on 8th March 1919. 

The editor, a former pupil of Dr. Moulton at Didsbury 
College, who had also worked under his guidance as a research 
student in Hellenistic Greek at Manchester University, was 
entrusted with the responsible duty of completing this volume 
and seeing it through the press. Part i. appeared in 1919, 
Part 11. in 1921. Meanwhile death had removed another 
worker whose tireless industry and unslumbering vigilance were 
well known to other toilers in this field. How sorely the editor 
has missed the help of Mr. Henry Scott may be gauged by 
comparing the number of misprints in Part ii. with the few 
corrections to be made in Part i., which owed so much to his 
careful reading of the proofs. Beyond writing the last thirteen 


vi PREFACE TO VOLUME II. 


pages of the Introduction, the editor’s responsibility in pre- 
paring the first two parts for publication was limited to the 
verification or insertion of numerous references and the less 
important though exacting labours of proof-correction. 

It is in the third part which now appears that the reader 
will recognise the immeasurable loss which this Grammar has 
suffered through the death of its brilliant author. Happily 
the chapter on Word-Composition can be printed almost exactly 
as it left the writer’s hands. Every student of New Testament 
lexicography will be grateful for this last gift of a great philo- 
logist. In passing to the chapter on Suffixes, the reader cannot 
fail to notice an abrupt change. No one is more acutely con- 
scious of the reader’s loss than the editor himself. In all 
matters of comparative philology, Dr. Moulton wrote with the 
authority of a master. The editor can only claim to have 
exercised the diligence of the scribe. He gladly acknowledges 
his debt to two writers in particular where many might be 
named. The late Karl Brugmann’s compendious Verglechende 
Grammatik has been in constant use, and his Griechische Gram- 
matik (edited by Albert Thumb in Iwan Miiller’s Handbuch der 
Klassischen Altertumswissenschaft) has been a close companion 
for years past. Grateful acknowledgment is also made to 
Professor Albert Debrunner, formerly of Bern, now of Jena, for 
his useful manual in Max Niedermann’s Sprachwissenschaftliche 
Gymnasialbibliothek, as well as for his contributions to the 
Indogermanische Forschungen. Other debts are freely acknow- 
ledged throughout the chapter. 

The discussion about the Semitic element in the Greek of the 
New Testament has passed into a new phase since Dr. Moulton 
projected his Appendix in conjunction with Mr. Bedale. This is 
chiefly due to three great Semitic scholars who have challenged 
the accepted theory regarding the original language in which 
the Acts, the Apocalypse, and the Fourth Gospel were written. 
Professor Torrey’s brilliant work on the Composition and Date 
of Acts appeared in 1916, but the editor first met with it while 


PREFACE TO VOLUME II. vil 


on a visit to America shortly after the war, when the first part 
of the Grammar had already passed through the press. Arch- 
deacon Charles had already impressed Dr. Moulton by some of 
his arguments in Studies in the Apocalypse, but the exhaustive 
examination of the grammar of Revelation came before the 
public with the issue cf the International Critical Commentary 
upon that book in the autumn of 1920. The lamented Professor 
Burney’s Aramaic Origin of the Fourth Gospel appeared in the 
summer of 1922. These books have aroused learned discussion 
among both Hellenists and Semitists, and the time has come for 
a critical survey of this entire field in its bearing upon the 
Grammar of New Testament Greek. If some readers are dis- 
posed to lament the long delay in completing the publication 
of this volume, others will be thankful that it has been possible 
to take full account of the most important literature since 
Wellhausen’s Einleitung, including the revised edition of Rader- 
macher’s Neutestamentliche Grammatik and the valuable linguistic 
studies by the eminent Semitist Pére Lagrange, in his Commen- 
taries on Luke (1921), Matthew (1923), and John (1925). 

Another feature in the Appendix deserves special mention. 
When the MS had already gone to the publisher in the spring 
of 1927 the editor had the good fortune to read a thesis by 
Dr. R. McKinlay, dealing with Semitisms in the New Testa- 
ment in the light of later popular Greek. It is to be hoped 
that this valuable work will soon find a publisher. Meanwhile, 
by the kindness of the author, the editor has been allowed 
to insert within square brackets an allusion to this work 
wherever Dr. McKinlay has proved that an alleged Semitism 
is an established construction in either Medieval or Modern 
Greek. The actual evidence will be forthcoming when the 
thesis is published. 

A word may be permitted with regard to the Indices. Limits 
of space forbid the registering of every Greek word that occurs 
in this volume. But prepositional compounds will generally 
be found by consulting the index under the prepositions, and 


CONTENTS. 


—— 
PAGE 

INTRODUCTION 5 1 

PART i. SOUNDS AND WRITING. 

THE ALPHABET, : = : : : : os 
Classification of Sounds 2 A 3 . . . 40 
Pronunciation . : ᾿ ; ; ς : ae 

MopeErRN PRINTED GREEK. Punctuation 6 5 ‘ . 46 

SOUNDS AND ORTHOGRAPHY— 

Accentuation . ὃ : 7 : : j . δ] 
Syllabification . : : : - : : . 60 
Klision , i : 5 i 3 5 : ol 
Crasis  . : : : : : : . 63 
Vowels . : : ; : ς : . 64 
Prothesis : : 3 ; : : : . 89 
Contraction ‘ ; : 5 : : i . 5389 
Syncope . : ᾿ : : : mos 
Pronunciation of the Vawelé : : : : ‘ . 98 
Aspiration ὁ 6 ὃ : : : - 397 
Single and Double ΤΌ. : : ; : . 101 
Intrusive Consonants . ‘ ‘ : : ; . 103 
Liquids and Nasals__. 5 : : ὃ : . 108 
Sibilants . : : ὃ : : : ‘ . 106 
Gutturals j Σ : a : : : . 108 
Labials . 3 : : : : ; : . 109 
Dentals . : ; ᾿ 3 7 ἢ Ἂν 5 ΝΠ) 
Movable Letters : : : : : ς 3 112 
PART ii. ACCIDENCE. 

DEFINITE ARTICLE ‘ : 5 : - : . 117 

Nouns— 

First Declension : ᾿ : : : Been lly) 


Second Declension ‘ 3 Ἶ : : ; - 120 


xi 


Xl CONTENTS. 


Metaplasmus and Heteroclisis 

Third Declension 

Declension of Semitic Names . 

Declension of Non-Semitic Names ; 
Miscellaneous Foreign Words and Irregular Ἤν 
Latin Nouns 


ADJECTIVES 


ADVERBS 
Comparison of Adjestives τ Adver Σ 


NUMERALS . 
PRONOUNS . 
VERBS é : 
Conjugation and Tense Stems 
Augment and Reduplication 
A. Present Stems— 
(a) Thematic— 
Active Voice 
Middle and Passive Tacs 
(6) Unthematic— 
Active Voice 
Middle and Passive Voices 
B. Strong Aorist Stem— 
Active Voice 
Middle Voice 
Passive Voice 
C. Weak Aorist Stem— 
Active Voice 
Middle Voice 
Passive Voice 
D. Future Stem— 
Active and Middle Voices 
Passive Voice 
E. Perfect Stem— 
Active Voice 
Middle and Passive Viens 


F. Verbals 


List or VERBS 


PAGE 


124 
128 
143 
150 
152 
155 


155 


163 
164 


167 
178 


182 
183 
188 


193 
197 


201 
205 


208 
212 
213 


215 
216 
217 


218 
219 


220 
223 


224 
224 


CONTENTS. 


PART iii. WORD-FORMATION. 


WorpD-CoMPOSITION 
Nouns. Classification— 

I. Co-ordinative Compounds . 
II. Iteratives 

IIt. Dependent αν θα 

IV. Descriptive Determinatives 
V. Possessive Compounds 

VI. Verbal Compounds . 


VII. Compounds based on Ere pactitiual eee 


PREPOSITIONS— 
I, Praeverbia 
II. Proper Prepositions 


Wee AN aus 2. Ἄνα; 9: ᾿ἄντύὸ; Ἔ ἀπὸ; 


Ὁ εἰς; We Ἐν; 8: ἘΡ; Ὁ: Ἡπί; 


1]. Μετά; 12. Παρά; 18. Περί; 


IIpos; 16. Suv; 17. ‘Yrép; 18. Ὑπό. 


III. Improper Prepositions . 


WorpbD-FORMATION BY SUFFIXES— 
Nouns anpD ADJECTIVES 
Formative Suffixes 
. Vowel Suffixes . 
. Nasal Suffixes . 
. Liquid Suffixes . 
. Suffixes with Labial Stops 
. Suffixes with Dental Stops 
. Suffixes with Guttural Stops . 
. Stems in -σ- 


ResBsoQdwpr 


VERBS— 
Class I. Person suffixes added to the root 

» II. With formative suffix in -n- 
» III. With formative suffix in so : se 
» ΙΝ. Suffixes in sko : ske 
59 V. Suffixes in fo : te 
» WI. Suffixes in do: de 
» VII. Suffixes in yo : ye 


i. Vocalic yo- presents 


Verbs in 1. -dw; 2. -€w; 3. 


δ. -vw; 6. -evo. 


5. Ata; 
10. 
14, Πρό; 


Κατά; 


15. 


ΧΙ] 


PAGE 


268 


269 
270 
271 
276 
283 
288 
291 


293 
294 


X1V CONTENTS. 


ii. Consonantal yo- presents 
A. Verbs in 1. -aivw; 2. -ύνω ἢ 
B. Verbs in 1. -aipw; 2. -cipw; 3. -ύρω 
C. Verbs in 1. -rrw; 2. -σσω 
D. Verbs in - ζω : ; : 
1. -d(w; 2. -ifw; 3. “ite 3 ; 4. -ἐζω ; 5. -ὀζω. 
APPENDIX—Semitisms ty THE NEw TESTAMENT . 


INDEX I.—Qvorations. (a) New Testament 
(6) Old Testament 
(c) Apocrypha and Peendlepierapin 

of Old Testament 
(4) Early Christian Writings . 
(e) Inscriptions and Ostraca . 
(f) Papyri 
(g) Greek Literature 
(A) Latin 
ἘΞ IIl.—Worps anp Forms. (a) Greek 
(6) Modern Greek 


(c) Hebrew and Aramaic 
Il].—Sussxrots . 


PAGE 


400 
401 
402 
403 
404 


41] 


487 
501 


502 
502 
503 
506 
512 
516 


517 
533 
534 
536 


ABBREVIATIONS. 


= 


Tur abbreviations for papyri and inscriptions are given in Index I (e) 
and (f), pp. 503-512 below, with the full titles of the collections quoted. 


References are to pages, unless otherwise stated. 


Abbott JV =Johannine Vocabulary, by Εἰ. A. Abbott. London, 1905. 
AJP=American Journal of Philology. Baltimore, 1880 ff. 
AJT=American Journal of Theology. Chicago, 1897 ff. 

Allen Comm. Matt.=A Critical and Exegetical Commentary on the Gospel 
acc. to St Matthew, by Willoughby C. Allen. 3rd ed. Edinburgh, 
1912. (See ICC.) 

Allen Comm. Mark=The Gospel acc. to St Mark, with Introduction and 
Notes, by Willoughby C. Allen. (The Oxford Church Biblical Com- 
mentary), London, 1915. 

Archiv—see Index I (e). 

Audollent—see Index I (e). 

Bauer HNT=Das Johannesevangelium erklart von Walter Bauer. 2te 
Aufl. Tiibingen, 1925. (See HNT.) 

Bauer Lex.=Griechisch-Deutsches Worterbuch zu den Schriften des Neuen 
Testaments und der tibrigen urchristlichen Literatur, von Walter 
Bauer. Giessen, 1924-28. 

Beginnings of Christianity=The Beginnings of Christianity, Pt. I., The 
Acts of the Apostles. Edited by F. J. Foakes Jackson and Kirsopp 
Lake. London, vol. i, 1920; vol. ii, 1922; vol. ii, 1926. 

Berl. Phil. Woch.=Berliner Philologische Wochenschrift. 

BIL-D, or Blass-Debrunner= Friedrich Blass’ Grammatik des neutestament- 
lichen Griechisch, bearbeitet von A. Debrunner. Gdéttingen, 4te Aufl. 
1913, 5te 1921. 

Blass Gr.=Grammar of NT Greek, by F. Blass. English tr. by H. St J. 
Thackeray. 2nded. London, 1905. 

Blass Philology=Philology of the Gospels, by F. Blass. London, 1898. 

Blass Pron.=Pronunciation of Ancient Greek. English tr. by Parton, 
1890. 

Boisacq = Dictionaire Etymologique de la Langue Grecque, par Emile 
Boisacq. Heidelberg and Paris, 1907-16. 

Bonhéffer=Hpiktet und das Neue Testament, von Adolf Bonhdoffer. 
Giessen, 1911. 

Bornhiuser= Die Bergpredigt. Versuch einer zeitgendssischen Auslegung, 
von Karl Bornhauser. Giitersloh, 1923. 

Bousset—see SNT7', 


XV 


XVi ABBREVIATIONS. 


Brugmann Dem.=Die Demonstrativpronomina der indogerm. Sprachen, 
von K. Brugmann. Leipzig, 1904. 

Brugmann Dist.=Die distributiven αι. d. kollektiven Numeralia der idg. 
Sprachen, von K. Brugmann. Leipzig, 1907. 

Brugmann Gr.* or Brugmann-Thumb=Giriechische Grammatik, von Karl 
Brugmann. 4te vermehrte Aufl., von Albert Thumb. Miinchen, 
1913. 

Brugmann Grd.2=Grundriss der vergleichenden Grammatik der indogerm. 
Sprachen, von K. Brugmann u. B. Delbriick. Strassburg. 2te 
Aufl. 1., 1897; 1. i, 1906; ii, 1911; iii, 1913-16. (For Syntax see 
under Delbriick.) 

Brugmann KVG=Kurze vergleichende Grammatik der idg. Sprachen, von 
Karl Brugmann. Strassburg, 1904. 

Buck Gr. Dial.=Introduction to the Study of the Greek Dialects, by C. D. 
Buck. Boston, 1910. 

Burkitt Hv. d. M.=Evangelion da-Mepharreshe, collected and arranged 
by F. C. Burkitt. Cambridge, 1904. 

Burkitt Gosp. Hist.=The Gospel History and its Transmission, by F. C. 
Burkitt. 3rd ed. Edinburgh, 1911. 

Burkitt Syr. Forms=Syriac Forms of NT Proper Names, by F. C. Burkitt. 
London, 1912. 

Burney Aram. Orig.=The Aramaic Origin of the Fourth Gospel, by C. F. 
Burney. Oxford, 1922. 

Burney Poetry=The Poetry of our Lord. An Examination of the Formal 
Elements of Hebrew Poetry in the Discourses of Jesus Christ, by 
C. F. Burney. Oxford, 1925. 

Burton Gal.=A Critical and Exegetical Commentary on the Epistle to the 
Galatians, by E. De Witt Burton. Edinburgh, 1921. (See JCC.) 

Cadbury AJ7'=Luke—Translator or Author ? by H. J. Cadbury. (Re- 
print from AJT' (see above), xxiv, No. 3, July 1920.) 

Cagnat—see Index I (e). 

Capes Ach. L.=The History of the Achaean League as contained in the 
remains of Polybius, edited with introduction and notes by W. W. 
Capes. London, 1888. 

CBE=Cambridge Biblical Essays, ed. H. B. Swete. London, 1909. 

CGT =Cambridge Greek Testament. 

Chandler=A Practical Introduction to Greek Accentuation, by H. Chandler. 
2nded. Oxford, 1881. 

Charles Asc. [sai.=The Ascension of Isaiah, translated from the Ethiopic 
Version, by R. H. Charles. London, 1900. 

Charles Hnoch=The Book of Enoch, ed. by R. H. Charles. Oxford, 1893. 

Charles Revelation=A Critical and Exegetical Commentary on the Revela- 
tion of St John, by R. H. Charles. 2 vols. Edinburgh, 1920. 
(See ICC.) 

Charles Studies=Studies in the Apocalypse, by R. H. Charles. Edinburgh, 
1913. 

Charles Test. XII Patr.=The Greek Versions of the Testaments of the 
Twelve Patriarchs, ed. by R. Η, Charles. Oxford, 1908. 


ABBREVIATIONS. XVII 


Charles—see also under OA. 

Chrest.—see Index I (f). 

Cl. Phil. =Classical Philology. Chicago. 

Cobet NT’ Vatic.=Novum Testamentum Graece ad fidem codicis Vaticanr 
restitutum, ab A. Kueno et C. G. Cobeto, cum praefatione Cobeti. 
Leiden, 1860. 

Conybeare and Stock Sel.—WSelections from the Septuagint, by F. C. 
Conybeare and St G. Stock. Boston, 1905. 

CQ—Classical Quarterly. London, 1907 ff. 

CR=Classical Review (London, 1887 ff.). Especially reference is made 
to J. H. Moulton’s collection of forms and examples from the papyri 
in OR xv, 31-38 and 434442 (Feb. and Dec. 1901), and xviii, 106-112 
and 151-155 (March and April 1904). 

Crénert Mem.= Memoria Graeca Herculanensis, by W. Cronert. Leipzig, 
1903. 

DAC=Dictionary of the Apostolic Church, ed. by J. Hastings. 2 vols. 
Edinburgh, 1915, 1918. 

Dalman Gr.2=Grammatik des Jiidisch-Paldstinischen Aramdisch, von 
Gustaf Dalman. 2te Aufl. Leipzig, 1905. 

Dalman WJ=The Words of Jesus, by G. Dalman. Eng. ed. tr. D. M. 
Kay. Edinburgh, 1902. 

Dalman Wérterbuch=Aramdische-neuhebradisches Handworterbuch zu 
Targum, Talmud und Midrasch, von Gustaf H. Dalman. 2te Aufl. 
Frankfurt a. Main, 1922. 

DB=Dictionary of the Bible, ed. by J. ae ae 5 vols. Edinburgh, 
1898-1904. 

Debrunner- Blass—see BI-D. 

Debrunner Worth.=Griechische Wortbildungslehre, von Albert Debrunner. 
Heidelberg, 1917. 

Deissmann BS=Bible Studies, by G. A. Deissmann. Eng, ed., including 
Bibelstudien and Neue Bibelstudien, tr. by A. Grieve. Edinburgh, 
1901. 

Deissmann In Christo=Die neutestamentliche Formel ‘in Christo Jesu, 
von G. A. Deissmann. Marburg, 1892. 

Deissmann LAE=Light from the Ancient Hast. Eng. tr. by L. R. M. 
Strachan. London, 1910. 2nd ed. (4th Germ. ed.), 1927. 

Deissmann Paul=St Paul, A Study in Social and Religious History. 
Eng. tr. by L. R. M. Strachan. London, 1912. 2nd ed., Eng. tr. 
(of 4th Germ. ed.) by W. E. Wilson, 1926. 

Delbriick= Vergleichende Grammatik der indogermanischen Sprachen, 
von K. Brugmann und B. Delbriick. Bde ITI-V, Syntax, von B. 
Delbriick. Strassburg (i) 1893, (ii) 1897, (iii) 1900. (Large Roman 
numerals indicate the volume number in the whole Grammar ; small 
Roman numerals in brackets give the volume in Delbriick’s Syntaz.) 

DLZ=Deutsche Literaturzeitung. Leipzig. 

G. R. Driver Orig. Lang.=The Original Language of the Fourth Gospel. 
A criticism of Dr Burney’s thesis, by G. R. Driver. (Reprinted 
from the Jewish Guardian, Jan. 5 and 12, 1923.) 


b 


”» 


XVill ABBREVIATIONS. 


S. R. Driver Tenses=A Treatise on the Use of the Tenses in Hebrew, by 
S. R. Driver. 3rd ed. Oxford, 1892. 

EBi=Encyclopedia Biblica, ed. by T. Κα. Cheyne and J. S. Blac. 
4 vols. London, 1899-1903. 

EGT = Expositor’s Greek Testament, ed. by ΝΥ. Robertson Nicoll. 5 vols. 
London, 1897-1910. 

Eranos=Eranos. Acta philologica Suecana. Upsala. 

Expos=The Expositor, ed. by W. R. Nicoll, afterwards by James 
Moffatt. London, 1875-1925. (Cited by series, volume and 
page.) 

Exp T=The Expository Times, ed. by J. Hastings, afterwards by A. W. 
and FE. Hastings. Edinburgh, 1889 ff. 

Meld Notes= Notes on the Translation of the New Testament, by Frederic 
Field. Cambridge, 1899. 

Fraenkel Geschichte d. Nom. Ag.=Geschichte der griechischen Nomina 
agentis auf -τήρ, -τωρ, -rns(-r-), von Ernst Fraenkel. Strassburg, 
1910, 1912. 

Fraenkel ZVS=(1) Zur Geschichte der Verbalnomina auf -σιο-, -cia. 
(2) Beitrige zur Geschichte der Adjectiva auf -τικός. Gottingen, 
1913. (For ZVS see below.) 

Ὁ and H—see P Oxy in Index I (/). 

Ges-K (or G-K)=Gesenius’ Hebrew Grammar, ed. by E. Kautzsch. Eng. 
tr. by Collins and Cowley. Oxford, 1910. 

Gildersleeve Studies=Studies in Honor of Basil L. Gildersleeve. Boston, 
1902. 

Giles Manual?=A Short Manual of Comparative Philology for Classical 
Students, by P. Giles. 2nd ed. London, 1901. 

Goodwin Gr. Gr?=A Greek Grammar, by ΝΥ. W. Goodwin. 2nd ed. 
London, 1894. 

Goodwin MT'=Syntax of the Moods and Tenses of the Greek Verb, by 
W. W. Goodwin. 3rd ed. London, 1889. 

Gregory Prol.—see under Ti. 

Guillemard=Hebraisms in the Greek Testament, by W. H. Guillemard. 
Cambridge, 1879. 

Harnack Luke=Luke the Physician, by A. Harnack. Eng. tr. by J. R. 
Wilkinson. London, 1907. 

Harnack Spriiche=Spriiche und Reden Jesu, von A. Harnack. Leipzig, 
1907. 

Harnack Sayings=The Sayings of Jesus. Eng. tr. by J. R. Wilkinson. 
London, 1908. 

Harris Codex Bezae=A Study of Codex Bezae, by J. Rendel Harris. 
Cambridge, 1891. 

Harris ΟΡ. =The Origin of the Prologue to St John’s Gospel, by Rendel 
Harris. Cambridge, 1917. 

Harris Testimonies=Testimonies, by Rendel Harris, with the assistance 
of Vacher Burch. Cambridge, pt. i, 1916; pt. 11, 1920. 

Harsing= De Optativi in Chartis Aegyptiis Usu, by Carl Harsing. Bonn, 
1910. 


ABBREVIATIONS. XIX 


Hatzidakis=FHinleitung in die neugriechische Grammatik, von G. N. 
Hatzidakis. Leipzig, 1892. 

Haupt=Die Gefangenschaftsbriefe (Meyers Kommentar ii. d. NT), von 
Erich Haupt. Gdéttingen, 1902. 

Hawkins HS=Hore Synoptice, by J. C. Hawkins. 2nd ed. London, 
1909. 

Heinrici= Die Korintherbriefe (Meyers Kommentar), von G. Heinrici. 
Gottingen, 1896. 

Heinrici Studien=Neutestamentliche Studien Georg Heinrici dargebracht. 
Leipzig, 1914. 

Helbing Gr.=Grammatik der Septuaginta : Laut- und Wortlehre, von R. 
Helbing. Gottingen, 1907. 

Herwerden Lex.= Lexicon Graecum suppletorium et dialecticum. 2nd ed., 
2 vols. Leiden, 1910. 

Hirt Gram. or Hbd.=Handbuch der Griechischen Laut- und Formenlehre, 
von Herman Hirt. Heidelberg, 1902. 

Hobart=The Medical Language of St Luke, by W. Καὶ. Hobart. Dublin, 
1882. 

HNT=Handbuch zum Neuen Testament, herausgegeben von Hans 
Leitzmann. Tibingen, 1907. 

H-R=A Concordance to the Septuagint, by EK. Hatch and H. A. Redpath. 
Oxford, 1897. 

AT R=Harvard Theological Review. Cambridge, Mass., 1908 ff. 

ICC=The International Critical Commentary. Edinburgh. 

IF or Idg F=Indogermanische Forschungen, ed. formerly by K. Brug- 
mann and ΝΥ. Streitberg. Strassburg, 1892 ff. 

IG, IGSI, IM Ae—see Index I (e). 

James Lang. of Pal.=The Language of Palestine and Adjacent Regions, 
by J. Courtney James. Edinburgh, 1920. 

Jannaris Gr.=A Historical Greek Grammar, by A. N. Jannaris. London, 
1897. 

JBL=Journal of Biblical Literature. New Haven, Conn. 

J HS—see Index I (e). 

Johannessohn Kasus τι. Prép.=Der Gebrauch der Kasus und der Pré- 
positionen in der Septuaginta, von Martin Johannessohn. Teil i. 
Berlin, 1910. 

JTS=Journal of Theological Studies. Oxford, 1900 ff. 

Kaibel—see Index I (e). 

Kautzsch Gr.=Grammatik des Biblisch-Aramdischen, von E. Kautzsch. 
Leipzig, 1884. 

K. Bl. and K-G—see Kiihner. 

Kennedy EG7T=The Epistle to the Philippians, ed. by H. A. A. Kennedy 
in the Exp. Greek Testament, vol. iii. London, 1903. 

Kennedy Sources=Sources of NT Greek, by H. A. A. Kennedy. Edin- 
burgh, 1895. 

Kieckers—see p. 9 n.}. 

Knowling=The Acts of the Apostles, ed. by R. J. Knowling in the Exp. 
Greek Testament, vol. ii. London, 1900. 


x ABBREVIATIONS. 


Kretschmer Hinl.=Die Einleitung in die Geschichte der griechischen 
Sprache, von P. Kretschmer. Gd6ttingen, 1896. 

Kretschmer Entstehung=Die Entstehung der Koine (Sitzungsberichte d. 
Wien. Akad., 1900). 

Kretschmer Vaseninschriften=Die griech. Vaseninschriften threr Sprache 
nach untersucht. Giitersloh, 1894. 

Kiihner or K.Bl., K-G=Ausfiihrliche Grammatik der griechischen Sprache, 
von R. Kiihner. 3te Aufl., Elementar- und Formenlehre, von F. 
Blass, i, ii. Hannover, 1890-92. Satzlehre, von B. Gerth, i, ii, 
1898, 1904. 

Kuhring=De Praepositionum Graec. in Chartis Aegyptiis Usu, by W. 
Kuhring. Bonn, 1906. 

KZ=Kuhns Zeitschrift fiir vergleichende Sprachforschung. Berlin and 
Giitersloh, 1852 ff. 

Lagrange S. Matt.=Evangile selon Saint Matthiew (Etudes Bibliques), 
par Le P. M.-J. Lagrange. Paris, 1923. 

Lagrange S. Marc=Evangile selon Saint Mare (Et. Bibl.). Paris, 
1920. 

Lagrange S. Luc=Evangile selon Saint Luc (Et. Bibl.). Paris, 1921. 

Lagrange S. Jean=Evangile selon Saint Jean (Et. Bibl.). Paris, 1925. 

Lake Cod. Sin.=Codex Sinaiticus Petropolitanus. With Introduction by 
Kirsopp Lake. Oxford, 1911. 

Law=The Tests of Life. A Study of the First Epistle of St John, by 
R. Law. Edinburgh, 1909. 

Lewy Fremdwérter=Die Semitischen Fremdworter im Griechischen, von 
H. Lewy. Berlin, 1895. 

Lietzmann HNT=Handbuch zum NT, iii. 1. Romerbrief, I. u. I. 
Korinther-, Galaterbrief, erklart von Hans Lietzmann. Tiibingen, 
1906 ff. 

J. Lightfoot Hor. Hebr.=Horae Hebraicae et Talmudicae, by John Light- 
foot (1658). Oxford, 1859. 

J. B. Lightfoot Ep. Ign. ad Polyc.=The Apostolic Fathers, by J. B. 
Lightfoot, part ii, 3 vols. 2nd ed. London, 1889. 

Lightfoot Fresh Revision=On a Fresh Revision of the English New Testa- 
ment, by J. B. Lightfoot. London, 1891. 

Lightfoot Notes=Notes on Epistles of St Paul from Unpublished Com- 
mentaries, by J. B. Lightfoot. London, 1895. 

Lindsay=The Latin Language, by W. M. Lindsay. Oxford, 1894. 

Lipsius Gr. Unt.=Grammatische Untersuchungen tiber die bibl. Grdacitat, 
von K. H. A. Lipsius. Leipzig, 1863. 

Lob. Paral.=Paralipomena Grammaticae Graecae, by C. A. Lobeck. 
Leipzig, 1837. 

Lob. Phryn.=Phrynichi Ecloga, ed. C. A. Lobeck. Leipzig, 1820. 

Lohmeyer HNT =Handbuch z. NT, iv. 1. Die Offenbarung des Johannes, 
erklart von Ernst Lohmeyer. Tiibingen, 1926. 

LS=A Greek-English Lexicon, by H. G. Liddell and R. Scott. 8th ed: 
Oxford, 1901. (The new LS=a new edition, revised and augmented 
throughout by Henry Stuart Jones. Oxford, 1925.) 


ABBREVIATIONS. ΣΙ 


Margolis Gr.=A Manual of the Aramaic Language of the Babylonian 
Talmud. Grammar, Chrestomathy and Glossaries, by Max L. 
Margolis. Miinchen, 1910. 

Marti Gr. (or K. Gr.)=Kurzgefasste Grammatik der Biblisch- Aramdischen 
Sprache, von Karl Marti. 4te Aufl. Berlin, 1911. 

Mayor Comm. James=The Epistle of St James. The Greek text with 
introduction, notes and comments, by J. B. Mayor. 3rd ed. 
London, 1910. 

Mayor Comm. Jude and 2 Pet.=The Epistle of St Jude and the Second 
Epistle of St Peter. Greek text with introduction, notes and com- 
ments, by J. B. Mayor. London, 1907. 

Mayser Gr.=Grammatik der gr. Papyri aus der Ptolemderzeit, von KE. 
Mayser. Leipzig, i. 1906; ii. 1. Berlin u. Leipzig, 1926. 

McKinlay =Semitisms in the New Testament in the Light of Later Popular 
Greek. Unpublished thesis by R. McKinlay. 

MecNeile Comm. Matt.=The Gospel acc. to St Matthew. The Greek text 
with introduction, notes and indices by A. H. McNeile. London, 
1915. 

Meisterhans*=Grammatik der attischen Inschriften, von Κα. Meisterhans. 
3te Aufl. von EK. Schwyzer. Berlin, 1900. 

Meélanges Nicole—see Index I (e). 

Melcher= De Sermone Epicteteo quibus rebus ab Attica regula discedat, by 
P. Melcher. Halle, 1906. 

Merx—Die vier kanonischen Evangelien nach threm diltesten bekannten 
Texte, von Adalbert Merx. Berlin, 1902-11. 

Meyer Gr.=Griechische Grammatik, von Gustav Meyer. 3te Aufl. 
Leipzig, 1896. 

Milligan Selections—see Index I (f). 

Milligan T'hess.=St Paul's Epistles to the Thessalonians, the Greek text, 
with introduction and notes, by G. Milligan. London, 1908. 

Mitteis Papyruskunde.=Grundzige und Chrestomathie der Papyruskunde, 
π΄. i., ed. L. Mitteis. See under Wilcken, also Index 1 (ἢ). 

Moeris= Moeridis Lexicon Atticum, ed. J. Pierson. Leiden, 1759. 

Moffatt ICC=A Critical and Exegetical Commentary on the Epistle to the 
Hebrews, by James Moffatt. Edinburgh, 1924. 

Moffatt Introd.=Introduction to the Literature of the NT. Edinburgh, 
ed.? 1912, ed.? 1918. 

Moffatt NT =A New Translation of the NT. 3rd ed. London, 1914. 

Monro Hom. Gr.=A Grammar of the Homeric Dialect, by 1). B. Monro. 
2nd ed. Oxford, 1891. 

Moulton Christian Religion=The Christian Religion in the Study and the 
Street, by J. H. Moulton. London, 1918. 

Moulton Hinl.=Einleitung in die Sprache des Neuen Testaments. (‘Trans- 
lated, with additions, from Proleg.*) Heidelberg, 1911. 

Moulton Prol. or Proleg®=A Grammar of NT Greek, by James Hope 
Moulton. Vol. i, Prolegomena. 3rd ed. Edinburgh, 1908. 

Nachmanson Beitr.—Beitriige zur Kenntnis der altgriechischen Volks- 
sprache, von E. Nachmanson. Upsala, 1910. 


ΧΧΙΙ ABBREVIATIONS. 


Nachmanson Magn.=Laute und Formen der Magnetischen Inschriften, 
von EK. Nachmanson. Upsala, 1903. 

Naigeli=Der Wortschatz des Apostels Paulus. Beitrag zur sprachge- 
schichtlichen Erforschung des Neuen Testaments, von Theodor 
Nageli. Gd6ttingen, 1905. 

Noldeke Beitr.=Beitrdge z. semit. Sprachwissenschaft, von Theodor 
Noldeke. Strassburg, 1904. 

Noldeke Syr. Gr?2=Kurzgefasste syrische Grammatik, von Theodor 
Néldeke. 216 Aufl. Leipzig, 1898. 

Norden Agnostos Theos=Untersuchungen zur Formengschichte religidser 
Rede, von Eduard Norden. Leipzig, 1913. 

Norden Ant. Kunstprosa= Die antike Kunstprosa vom VI Jahrhundert νυ. 
Chr. bis in die Zeit der Renaissance, von E. Norden. 2 Bde. Leipzig, 
1915. 

O(aford) A(pocrypha)=The Apocrypha and Pseudipigrapha of the Old 
Testament in English, ed., in conjunction with many scholars, by 
R. H. Charles. 2 vols. Oxford, 1913. 

OCT (Oxford Classical Texts)=Seriptorum Classicorum Bibliotheca 
Oxoniensis. 

OGIS—see Index I (e). 

OHL (or Oxf. Heb. Lex.)=Hebrew and English Lexicon of the OT, ed. by 
F. Brown, 8. R. Driver, and C. A. Briggs. Oxford, 1906. 

Ottley Isaiah=The Book of Isaiah acc. to the Septuagint, trans. and ed. 
by R. R. Ottley. 2 vols. 2nd ed. Cambridge, 1909. 

Oxford Studies=Oxford Studies in the Synoptic Problem, ed. by W. Sanday. 
Oxford, 1911. 

Parry=The Pastoral Epistles, with introduction, text and commentary, 
by R. St J. Parry. Cambridge, 1920. 

Peake’s Commentary=A Commentary on the Bible, ed. by Arthur 8. Peake. 
Edinburgh, 1919. 

Pelagia=Legenden der heiligen Pelagia, ed. H. Usener. Bonn, 1879. 

Petersen Gr. Dim.=Greek Diminutives in -ION. A Study in Semantics, 
by Walter Petersen. Weimar, 1910. 

v. d. Pfordten=Zur Geschichte der griechischen Denominativa, von H. F. 
von der Pfordten. Leipzig, 1886. 

Preisigke Ostr. and Sammelb.—see Index I (e). 

Prellwitz—Etymologisches W6rterbuch der griechischen Sprache, von 
Walther Prellwitz. Gé6ttingen, 1892. 

Psichari—Essai sur le Grec de la Septante, par Jean Psichari. (Extrait 
de la Revue des Etudes juives, Avril 1908.) Paris, 1908. 

Radermacher Gr.=—Neutestamentliche Grammatik (HNT i. 1), von L. 
Radermacher. Tiibingen, 1911. 2te Aufl., 1925. 

Ramsay C’. and B.—see Index I (6). 

Ramsay Paul=Paul the Traveller and Roman Citizen, by W. M. Ramsay. 
3rd ed. London, 1897. 

Ramsay Teaching of Pawl=The Teaching of Paul in the Terms of the 
Present Day. London, 1915. 

REGr= Revue des Etudes grecques. Paris, 1888 ff. 


ABBREVIATIONS. XXiil 


Reinhold =De Graecitate Patrum, by H. Reinhold. Halle, 1898. 

Riddell=A Digest of Platonic Idioms, by J. Riddell (in his edition of the 
Apology). Oxford, 1867. 

R. McK.—see under McKinlay. 

Roberts =Introduction to Greek Epigraphy, by E. 8. Roberts. Cambridge, 
1887. 

Robertson Gr.=Grammar of the Greek Testament in the Light of Historicat 
Research, by A. T. Robertson. New York, 1914. 3rd ed., 1919. 
Rossberg=De Praepositionum Graecarum in Chartis Aegyptiis Ptole- 

maeorum Aetatis Usu. C. Rossberg. Jena, 1909. 

Rouffiac = Recherches sur les charactéres du grec dans le Nouveau Testament 
@aprés les inscriptions de Priéne, par Jean Rouffiac. Paris, 1911. 

Rudberg=Neutestamentliche Text und Nomina Sacra, Upsala, 1915. 

Rutherford, Gram.=First Greek Grammar, by W. G. Rutherford. 
London, 1907. 

Rutherford NP=The New Phrynichus, by W. G. Rutherford. London, 
1881. 

Sanders=Facsimile of the Washington Manuscript of the Four Gospels 
in the Freer Collection, with an introduction by H. A. Sanders. 
Michigan, 1912. 

Sandys and Paley=Select Private Orations of Demosthenes. Cambridge, 
1875. 3rd ed., 1898. 

Scham=Der Optativgebrauch bei Clemens von Alexandrien, von Jakob 
Scham. Paderborn, 1913. 

Schmid Aftic.=Der Atticismus in seinen Hauptvertretern von Dionysius 
von Halikarnass bis auf den zweiten Philostratus, von W. Schmid. 
4 Bde und Register. Stuttgart, 1887-97. 

Schmidt Jos.=De Flavii Josephi Elocutione, by W. Schmidt. Leipzig, 
1893. 

Schrade=Reallexikon der indogermanischen Altertumskunde, von O. 
Schrade. Strassburg, 1901. 

Schubart=Linfiihrung in die Papyruskunde, von W. Schubart. Berlin, 
1918. 

Schulthess Das Problem= Das Problem der Sprache Jesu, von F. Schulthess. 
Ziirich, 1917. 

Schulthess ZVT'W (see below)=Zur Sprache der Evangelien. Giessen, 
1922. 

Schwyzer=Grammatik der pergamenischen Inschriften, von EK. Schweizer 
(since Schwyzer). Berlin, 1898. 

Serivener Codex Bez.=Bezae Codex Cantabrigiensis, ed., with a critical 
introduction, annotations and facsimiles, by F. H. Scrivener. 
Cambridge, 1864. 

Serivener Collation=A full Collation of the Codex Sinaiticus with the 

; Received Text of the NT, by F. H. Scrivener. Cambridge, 1864. 
2nd ed., 1867. 

SH=A Critical and Exegetical Commentary on the Epistle to the Romans 
(ICC), by W. Sanday and A. C. Headlam. 5th ed. Edinburgh, 
1902. 


XXIV ABBREVIATIONS. 


Sharp Epict.= Epictetus and the New Testament, by 1). 8. Sharp. London, 
1914. 

Simecox—The Writers of the New Testament, by W. H. Simcox. 2nd ed. 
London, 1902. 

Smith L and L of Paul=The Life and Letters of St Paul, by David Smith. 
London, 1919. 

SNT=Die Schriften des Neuen Testaments, herausgegeben von Johannes 
Weiss. 3te Aufl. Gottingen, 1917. (Gal. 1 u. 2 Kor. erkl. von W. 
Bousset. ) 

v. Soden, Die Schriften=Die Schriften d. Neuen Testaments in thre 
diltesten erreichbaren Textgestalt, hergestellt von Hermann Freiherr 
von Soden. 4 Bde. Berlin, 1902-13. 

Solmsen Gr. Wortf.—Beitrdge zur Griechischen Wortforschung, von Felix 
Solmsen. Ite Teil. Strassburg, 1909. 

Sophocles Lex.=Greek Lexicon of the Roman and Byzantine Pertods, by 
E. A. Sophocles. Boston, 1870. 

Souter Lex.=A Pocket Lexicon to the Greek New Testament, by Alexander 
Souter. Oxford, 1917. 

Stahl Synt.—Kritisch-historische Syntax des griechischen Verbums der 
klassischen Zeit, von J. M. Stahl. Heidelberg, 1907. 

Stephanus Thes.=Thesaurus Graecae Linguae, ab Henrico Stephano 
constructus. 5 vols. Paris, 1572. 

Stevenson Gr.—Grammar cf Palestinian Jewish Aramaic, by W. B. 
Stevenson. Oxford, 1924. 

Siitterlin—Zur Geschichte der Verba Denominativa im Altgriechischen, 
lte Teil: die Verba Denominativa auf -dw -έω -όω, von Ludwig 
Siitterlin. Strassburg, 1891. 

Syll.—see Index I (e). 

Thackeray Gr.=A Grammar of the OT in Greek, i, by H. St J. Thackeray. 
Cambridge, 1909. 

Thackeray Schweich Lectures=The Septuagint and Jewish Worship. 
London, 1922. 

ThLZ—=Theologische Literaturzeitung. Leipzig, 1876 ff. 

Thompson Palaeography=Handbook to Greek and Latin Palaeography, by 
E. Maunde Thompson. 2nd ed. London, 1894 ; 3rd ed., 1913. 

Thumb—see Brugmann Gr.* 

Thumb Dial.=Handbuch der griechischen Dialekte, von A. Thumb. 
Heidelberg, 1909. 

Thumb Handb.=Handbook of the Modern Greek Vernacular, by A. Thumb. 
Translated from the second German ed. by 8. Angus. Edinburgh, 
1912. 

Thumb Hellen.—Die griechische Sprache im Zeitalter des Hellenismus, 
von A. Thumb. Strassburg, 1901. 

Thumb Sp. Asper=Untersuchungen tiber d. Spiritus Asper im Griech- 
ischen, von A. Thumb, Strassburg, 1889. 

Ti=Novum Testamentum Graece, by C. Tischendorf. Editio octava 
critica maior. 2 vols. Leipzig, 1869-72. Also vol. iii, by Ὁ. R. 
Gregory, containing Prolegomena, 1884. 


ABBREVIATIONS. XXV 


Torrey CDA=The Composition and Date of Acts, by C. C. Torrey. Camb. 
Mass., 1916. 

Torrey, C. H. Toy Studies=The Translations made from the original 
Aramaic Gospels, by C. C. Torrey. (See C. H. Toy Studies.) 

Torrey HT R=The Aramaic Origin of the Gospel of John. (See HT'R.) 

C. H. Toy Studies=Studies in the History of Religions presented to Crawford 
Howell Toy. New York, 1912. 

Valaori=Der delphische Dialekt, von J. Valaori. Gottingen, 1901. 

Veitch= Verbs Irregular and Defective, by W. Veitch. Oxford, 1887. 

Viereck SG—see Index I (e). 

Viteau Etude sur le grec du Nouveau Testament, par J. Viteau. i, Paris, 
1893 ; ii, 1896. 

Vocab.=The Vocabulary of the Greek Testament ilustrated from the Papyri 
and other non-literary sources, by J. H. Moulton and G. Milligan. 
London, 1914— 

Vogeser=Zur Sprache der griechischen Heiligenlegenden, von J. Vogeser. 
Miinchen, 1907. 

Volker=Papyrorum gr. Syntaxis Specimen, von F, Volker. Bonn, 1900. 

W1, W?—see Wellhausen. 

Wackernagel Hellen. =Hellenistica (Hinladung zur akadem. Preisverkiind- 
digung), von Jacob Wackernagel. Gottingen, 1907. 

Wackernagel Vorlesungen=Vorlesungen tiber Syntax. Erste Reihe. 2te 
Aufl. Basel, 1926; zweite Reihe, 1924. 

Walde Lat. Htym. Wort.—Lateinisches etymologisches Worterbuch, von 
A. Walde. Heidelberg, 1906. 2te Aufl., 1910. 

J. Weiss=Der erste Korintherbrief (Meyers Kommentar), von Johannes 
Weiss. G6ttingen, 1910. 

Wellhausen=FHinleitung in die drei ersten Evangelien, von J. Wellhausen. 
Berlin, 106 Aufl. (171), 1905; 2te Aufl. (W?), 1911. 

Wellhausen Das Evangelium Matthaei. Berlin, 1904. 

Wellhausen Das Hvangelium Marci. Berlin, 1903. 

Wellhausen Das EHvangeliwm Lucae. Berlin, 1904. 

Wellhausen Das Hvangelium Johannis. Berlin, 1908. 

Wendland=Die urchristlichen Literaturformen, von Paul Wendland. 
(HNT 1.3.) 2te Aufl. Tiibingen, 1912. 

Wernle=Die synoptische Frage, von Paul Wernle. Freiburg i. B., 
1899. 

Wessely Studien=Studien zur Palaeographie und Papyruskunde, heraus- 
geg. von C. Wessely. Leipzig, 1901. 

Wessely Zauberpap.—C. Wessely, Griechische Zauberpapyri von Paris und 
London. See Index I (f), under P Par 574. 

Wetstein=Novum Testamentum Graecum, by J. J. Wetstein. 2 vols. 
Amsterdam, 1751. 

WH=The New Testament in the Original Greek, by B. F. Westcott and 
F. J. A. Hort. Vol. i, Text (also ed. minor) ; vol. ii, Introduction. 
Cambridge and London, 1881 ; 2nd ed. of vol. ii, 1896. 

WH App=Appendix to WH, in vol. ii, containing Notes on Select 
Readings and on Orthography, etc. 


ΧΧΥῚ ABBREVIATIONS. 


Wilamowitz Lesebuch—Griechisches Lesebuch, von U. von Wilamowitz- 
Moellendorff. i. Text; ii. Erlauterungen. 3te unverand. Aufi.. 
1903 : 7te unverand. Aufi., 1920. Berlin. 

Wilcken Papyruskunde.—Grundziige und Chrestomathie der Papyruskunde, 
1. i, ed. U. Wilcken. See under Mitteis, also Index I (/). 

Wilcken—see Archiv and UPZ in Index I (e) and (f.) 

Windisch=Der zweite Korintherbrief (Meyers Kommentar), von H. 
Windisch. Gottingen, 1924. 

Witk.—-see Index I (ῇ. 

WM=A Treatise on the Grammar of New Testament Greek, regarded as a 
sure basis for NT Exegesis, by G. B. Winer. Translated from the 
German, with large additions and full indices, by W. F. Moulton. 
3rd ed. Edinburgh, 1882. 

Wright Comp. Gr. Gram.—Comparative Grammar of the Greek Language, 
by Joseph Wright. Oxford, 1912. 

WS=6. 8. Winers Grammatil: des neutestamentlichen Sprachidioms. Ste 
Aufi. neu bearbeitet von P. W. Schmiedel. Gottingen, i. Teil, 1894: 
ii. Teil, erstes Heft, 1897: zweites Heft, 1898. 

Zahn=Introduction to the NT, by Theodor Zahn. Eng. tr. of 3rd ed. 
Edinburgh, 3 vols.. 1909. 

Zahn Matthaus—Das Evangelium des Maitthaus, ausgelegt von Theodor 
Zahn. Leipzig. Ite Aufl. 1903, 4te Aufl. 1922. 

ZNT W=Zeitschrijft far die neutestamentliche Wissenschaft. Giessen 
1900 #. 

ZVS=Zeitschrijft far vergleichende Sprachforschung auf dem Gebiete der 
indogermanischen Sprachen. Begrundet von A. Kuhn. Gottingen. 


Signs : 
For 7 see pp. 225 and 334. 
For *, +, J see p. 292. 


ADDENDA ET CORRIGENDA. 


--ςφ 


Page 10, line 11, after Luke, insert (1238), 


39 


39 


16, line 14, read Aramaic. 

19, n.2 line 3 ab imo, read Lietzmann. 

22, line 4, read 1 Co 15*°. 

22, line 17, after Ephesians read (429, 5°). 

44, ὃ 25, inset title, wnder Consonants, insert Mutes. 

57, line 8, before Moeris insert See p. 209 no; 

70, line 20, after WS 47 f., insert also Charles Rev. (ICC) i. 216. 

82, line 2 ab imo, for indicio read iudicio. 

91, line 10, for ὃ 53 read ὃ 52. 

99, n.1 line 2 ab imo, first word, read that. 

100, line 6, for ὃ 76 read ὃ 77. 

103, line 13, read Grd.2 I. 827. 

103, line 18, after λήμψομαι insert (p. 106), after σφυδρόν (p. 112). 

121 C. (Ὁ), read Ὕλεως survives in NT only in the nom. sing (but 
see Clem ad Cor. passim). 

130, line 5 ab imo, for (=i. 109-88) read (=i. 166-88). 

131, line 10 for -ts read -ὕς. 

131, (3) (ὁ) line 2, read association. 

135 (b) line 4, read (=-n- σι). 

135, line 10 ab imo, read ὠδῖν. 

172 (b) Arrangement should be in italics. 

173 (4) line 2, for χιδιάδες read χιλιάδες. 

175, line 11, for τέσσαρις read τέσσερις. 

176, line 3, after Appendix, insert p. 439. 

176, line ὃ ab imo, Fractions should be in italics. 

177 (0) last line, read Tob 1010 Β. 

192 7. line 6, insert comma after (ἀπόλογος). 

192 8. line 2, znsert bracket ) after ulcus. 

192 n.°, line 2, read Jer 43(36)?°. 

199 9. line 3, read épp. 


198, line 8, read Hermas Vis. iii. 107. 
XXvii 


XXVill ADDENDA ET CORRIGENDA. 


Page 199, line 8 ab imo, read φιλώμεθα. 


39 


204, line 8, insert 3 under | and 2. 

207, line 18, for 7167 read τιθῇ. 

210. Subjunctive. In 2 sg. read δῷς, dois. 
219, line 16, after see insert p. 216. 

223, line 22, after Hermas insert Mand. xii. 
225, line 10 ab imo, for -οσα read -υσα. 


Or 
τῷ 
. 


226 aipéw, 3rd col. read ἡρέθην. 

231 βούλομαι, for (11. a) read (1. a). 

232, after δέομαι insert (I. a). 

234, *Fevd, last col., for 78ew read ἤδειν. 

235, *Fecx for present read perfect. 

235, ἐλαύνω Comp., for ἀπ- read ἀπ-. 

238, line 4, for ἐρωτήθω read ἐρωτήσω. 

239, after Cwypéw insert (VIL). 

242, line 15, after κάθημαι insert (For flexion see § 87.). 

242, line 20, after καθίζω insert (VII.). 

243, line 10 ab imo, after κεράννυμι for (1. 3) read (11. 8). 

246, line 3 ab imo, read -λημπτός ἡ. 

247, λείπω 3rd col., read ἐλείφθην 

257, line 6, for pyle read ῥήξω. 

259, line 12 ab imo, read turn. 

262, line 5 ab imo, for ἔφᾶνα read ἔφᾶνα. 

266, line 3 ab imo, for -éwoa read -ἔωσα. 

276. Good examples of K.D. cpds. are πρωτοπολίτης P Oxy 1. 414, 
and προβατοκτηνοτρόφοι P Ry] ii. 73° (33-30 B.c.) and note. 

470, line 9. Possibly we should add a third class (c) ἵνα for ὅτι 
(=). So Archd. Allen attempts to explain the hard saying in 
Mk 412 (Comm. Mark, p. 80). But it is simpler, with Lagrange, 
to take iva as equivalent to iva πληρωθῇ in introducing the 
citation. 


Dr. J. H. Moulton’s 


Grammar of New Testament Greek 


“marks a revolution in New Testament study.” 
Dr. J. HasTincs in Zhe Exposetory Times. 


VOL. I. THE PROLEGOMENA. 
Third Edition. Demy 8vo, cloth. 10/- net. 
CONTENTS :—General Characteristics—History of the 
‘““Common” Greek—Notes on the Accidence— 
Syntax: The Noun—Adjectives, Pronouns, Pre- 
positions—The Verb: Times and Modes of Action 
—Voice—Moods—The Infinitive and Participle— 
Additional Notes—Indices to Quotations, Greek 

Words and Forms, Subjects. 


VOL. II. ACCIDENCE AND WORD-FORMATION. 


Part i. General Introduction. 
Sounds and Writing. 


Just Published. In paper covers. 7/- net. 


Part ii. Accidence. 
(a) Nouns; (ὁ) Verbs and Lists. [/ Preparation. 


Part iii. Word-Formation. [lx Preparation. 


When completed, the Three Parts of Vol. II. 
may be bound together in cloth binding. 


EpinspurGH: T. & T. CLARK, 38 GEORGE STREET. 


A GRAMMAR OF 
NEW TESTAMENT GREEK 


At the Classroom Door, 


—— 


ORD, at Thy word opens yon door, inviting 
Teacher and taught to feast this hour with Thee ; 
Opens a Book where God in human writing 
Thinks His deep thoughts, and dead tongues live for me. 


Too dread the task, too great the duty calling, 
Too heavy far the weight is laid on me! 

O if mine own thought should on Thy words falling 
Mar the great message, and men hear not Thee! 


Give me Thy voice to speak, Thine ear to listen, 
Give me Thy mind to grasp Thy mystery ; 

So shall my heart throb, and my glad eyes glisten, 
Rapt with the wonders Thou dost show to me. 


JAMES HOPE MOULTON. 


BANGALORE, 
February 21, 1917. 


A GRAMMAR OF 
NEW TESTAMENT GREEK 


BY 


JAMES HOPE MOULTON 
M.A. (CANTAB.), D.Lit.(LoNnpD.), D.D.(EDIN.), D.C.L.(DUNELM.) 
Ὁ. THEOL. (BERLIN AND GRONINGEN) 


LATE FELLOW OF KING'S COLLEGE, CAMBRIDGE 
LATE GREENWOOD PROFESSOR OF HELLENISTIC GREEK AND INDO-EUROPEAN PHILOLOGY 
MANCHESTER UNIVERSITY 
LATE TUTOR IN NEW TESTAMENT LANGUAGE AND LITERATURE 
WESLEYAN COLLEGE, DIDSBURY 


Vol. Il 
ACCIDENCE AND WORD-FORMATION 


Part i 


GENERAL INTRODUCTION 
SOUNDS AND WRITING 


EDITED BY 


WILBERT FRANCIS HOWARD 
M.A.(Mancu.), B.D.(Lonp.) 


Eprnpurcu: T. & T. CLARK, 38 GerorGE STREET 
De ee, 


PRINTED BY 


Morrison & Giss LIMITED, 
FOR 


τ ἃ T. CLARK, EDINBURGH. 


LONDON : SIMPKIN, MARSHALL, HAMILTON, KENT, AND CO. LIMITED. 
NEW YORK: CHARLES SCRIBNER’S SONS. 


PREFACE. 


τεσ πες 


THE first volume of this Grammar of New Testament Greek 
was published in January 1906. Almost before the last 
sheets of the Prolegomena had passed from his hands Dr. 
Moulton was at work on the second volume. From that time 
until October 1915 when he set sail for India, the preparation 
of the second volume had first claim on his leisure hours and 
was also residuary legatee when the insistent demands of the 
pulpit, the platform and the committee room had been duly 
met. Three times only in those ten years was this task laid 
aside. In the autumn of 1911 he undertook a visitation of 
the Mission Stations of his Church in the West Indies on his 
way to the Cicumenical Methodist Conference at Toronto. 
The following year the delivery and publication of the 
Hibbert Lectures delayed his work on the Grammar. Α less 
serious interruption was a brief visit to the States, just before 
the outbreak of war, to give a series of popular lectures at 
Northfield. With these exceptions the work went steadily 
on, as the dated entries on the MS bear witness, though the 
issue of a fresh volume of papyri, or the appearance of some 
German monograph, or an event like the discovery and publica- 
tion of the Washington Codex, meant the re-examination and 
sometimes the revision of the portion already written. Ina 
letter to his publisher towards the end of this time he writes 
sadly that his college and university duties only leave him a 
few hours daily for the Grammar. 

At last, when the long toil was almost ended, sorrow 
invaded the home. The sudden death of his wife over- 
whelmed him. Then in his hour of grief a door of hope 
opened, and he responded to a call from the Y.M.C.A. to give 
a course of lectures on Zoroastrianism and Christianity to 


the Parsi community in Bombay. Eighteen months were 
vii 


Vill PREFACE. 


spent in India and Ceylon—months of breathless activity 
which brought healing though at the cost of much physical 
exhaustion. But there was still another heavy blow to fall. 
During his stay in India news reached him that his elder son 
had fallen in action in the Somme campaign late in the 
summer of 1916. Dr. Moulton was far from well when the 
City of Paris set sail from Karachi in the following March. 
The story of that voyage and of its tragic close has been told 
once for all by that dear friend who joined him at Alexandria 
and was with him to the end." 

Readers of the preface to the first edition of the Prolego- 
mena will not need to be reminded of the cruel irony that 
such a doom should fall upon this passionate lover of inter- 
national goodwill. He fell a victim to that frenzy of hate 
which broke with every sacred tradition of the sea. We 
know that he toiled like a hero, and kept a cheerful courage 
through the three days and two nights of winter storms in 
an open boat until he succumbed not only to bodily weak- 
ness, but to “superior spiritual attractions which he felt a 
long time before the ship was struck.” Who that knew him 
can doubt that in his heart also there went up the prayer, 
Κύριε, μὴ στήσῃς αὐτοῖς ταύτην τὴν ἁμαρτίαν. 

It has fallen to other hands to prepare the Grammar for 
publication and to see the sheets through the press. The 
second volume consists of three parts: (1) Sounds and writ- 
ing; (2) Accidence; (3) Word-formation. The first and 
second parts were complete; the third part was left unfinished. 
Shortly before leaving this country, however, Dr. Moulton 
decided upon an important addition to the original plan of 
Vol. Il. An introductory chapter was to follow up some 
questions raised in Vol. I. and to prepare the way for an 
Appendix dealing minutely with the vexed question of Semi- 
tisms in the New Testament. Writing to his publisher 
(September 7, 1915) he says, “The Introduction will not 
cover the ground of Prolegomena but will be supplementary, 
and will also include discussions of matters which have been 
in debate since the book appeared nine years ago. Chief 


1 See letter from Dr. Rendel Harris to the Rev. W. Fiddian Moulton in 
Bulletin of the John Rylands Library, Manchester, vol, iv. No. 1. 


PREFACE. ΙΧ 


among these is the question of Semitism, and I want to deal 
with this in a double way, which will make it one of the 
most conspicuous features of the book. The Introduction 
deals with it generally, but will refer to an Appendix which 
is to take up the whole matter in detail. There is nothing 
at present which tackles the question at all systematically 
according to modern lights. For that purpose I need to 
have the help of a skilled Semitist, and I have asked my 
colleague the Rev. C. L. Bedale, who takes Hebrew at Dids- 
bury and Assyrian at the University, to collaborate with me 
in this matter. He has very gladly consented, and he will 
be gathering material while 1 am away. He is an extremely 
sound and able scholar, and, of course, we have the great 
advantage of being neighbours and able to consult constantly. 
I propose to put his name on the title page for the Appendix 
in question. I am sure that it will add very considerably to 
the value of the book.” 

Dr. Moulton at once began to write the Introduction 
currente calamo, but some interruption disturbed him in the 
middle of a sentence when only two-thirds of the chapter 
had been written. The editor must, of course, assume sole 
responsibility for the remainder of the Introduction (pp. 22— 
34), but he believes that what he has supplied is a faithful 
representation of Dr. Moulton’s opinion on the questions under 
discussion. He was so fortunate as to track down a paper 
(written with great care only eighteen months earlier than 
the date of the Introduction) of which large use is made in 
§ 14 and from which a few sentences are quoted in later 
sections. Some further extracts from Dr. Moulton’s contribu- 
tion to Cambridge Biblical Essays (by kind permission of 
Messrs. Macmillan) and many jottings from Dr. Moulton’s 
College and University lectures made it possible to complete 
the chapter according to the author’s design. 

The difficulties in the way of printing and publication 
have been, and still are, very formidable. But for the 
generosity and enterprise of Messrs. Τὶ & T. Clark, the appear- 
ance of this book would have been delayed indefinitely. As 
it is, Vol. II. will be published in three separate parts in paper 
covers. The second part will follow closely on the heels of 
this portion, and the third part will be published with the least 


x PREFACE. 


possible delay. The editor is now at work on the chapter 
that will complete part iii. Mr. Bedale’s work on the 
Appendix has been interrupted for the past three years by 
service abroad as Chaplain to the Forces. 

_ In acknowledging obligations one thinks first of those 
friends of whom the author would have rejoiced to make 
grateful mention. Not even the war could have silenced his 
tribute to those two friends who had spent happy days as his 
guests at Didsbury not many months before the storm broke 
—Professor Albert Thumb, who died on August 14, 1915, 
and Professor Adolf Deissmann. His debt to many conti- 
nental scholars is sufficiently indicated in the footnotes. But 
personal gratitude would most certainly have been expressed 
to such friends as Dr. Rendel Harris, Mr. H. St. John 
Thackeray, Professor Souter, Dr. Rouse and other specialists 
with whom he was in constant correspondence. 

The editor returns most hearty thanks to Professor 
Milligan who has not only read the proof sheets with closest 
care, checking references to papyri, but has given the benefit 
of his counsel again and again when difficult points called for 
settlement; to Professor Peake for valuable information and 
advice in several matters; and not least to Mr. H. Scott, 
whose name is by this time so familiar to all who are interested 
in the minutiae of Greek Testament grammar. Not only has 
Mr. Scott gone through the proofs most carefully in all stages 
to their great gain in accuracy, but he is also preparing the 
Indexes for the whole of Vol. IL, and will fulfil a promise 
made to Dr. Moulton ten years ago to provide an Appendix 
elaborating certain principles that are lightly traced in 
part 11]. 

Something more than formal gratitude is due to Messrs. 
Morrison & Gibb for their share in the production of this 
book during the most difficult months ever known in the 


printing world. 
W. F. HOWARD. 


WALLASEY, CHESHIRE, 
Christmas Day, 1918, 


SYNOPSIS OF 
INTRODUCTORY CHAPTER. 


—  ---- 

i PAGE 

81. A Supplement to Prolegomena—New Testament 
Greek in the light of recent discussions . 5 1 

1. New TESTAMENT GREEK As A UNiITY— 

§ 2. It is in general the colloquial lingua franca of the 
Empire : - : : : : au; 

§ 3. First impulse to this use of the lingua cottidiana 
comes from LXX 2 

§ 4. Fitness of Hellenistic vernacular as a medium for 
expressing Christian gospel - . Ξ 9 

2. CoNTACTS WITH LitERARY LANGUAGE— 

§ 5. Atticism of 2 Pe. . : ἢ : - 5 
§ 6. Hebrews.—Complete absence of potential optative . 6 

Lucan Writings. Same test shows Luke as the 

only littérateur among NT writers. Variety of 
style according to changing scene of narrative. 7 

§ 7. Pauline Writings.—Use of vernacular inflexions as 

by Luke, but more popular vocabulary and less 
literary style : - : ; : 8 


§ 8. The First Gospel.—Treatment of sources: uni- 

formity of style, genius for compression, fondness 

for parallelism, substitution of literary for ver- 
nacular forms : : : : Be eal) 

Use by all these writers of distinctly Hellenistic 

flexions, avoidance of dual, final optative, and 

other hall-marks of Atticism, archaisms, and 
affectations . : Σ : : LD) 


a 


8 9. False test of compound verbs. WHarnack’serror . 1] 
xi 


xl SYNOPSIS OF INTRODUCTORY CHAPTER. 


3. SEMITIC CoLOURING— 


(1) Language Conditions of Palestine . 


§ 10. 


Difference between Jerusalem and Galilee. Gali- 
leans bilingual 


(2) Aramaisms and Hebraisms 


8.11: 


812 


§ 13. 
§ 14. 


Semitism in NT defined as “a deviation from 
genuine Greek idiom due to too literal rendering 
of the language of a Semitic original.” The 
resulting sense may be (1) identical or nearly 
identical. A literal rendering may produce a 
very marked overuse of a rare but permissible 
locution : 6.4. ἐνώπιον and ἰδού. Secondary Semi- 
tisms. If resulting sense is (2) different from 
that of original, need of caution in conjecturing 
Aramaic original 


True Semitisms in NT of two kinds 


(1) Imitations, conscious or unconscious, of LXX 
“translation-Greek ” : Hebraisms 


(2) Similarly slavish renderings of Semitic sources, 
oral or written: Aramaisms . 
N.B.—Strong influence of Pentatenel {πὸ 


Kown) . 

Extent to which Sk Pee τι πον οὶ ΠΣ 
occur in the several writers . : : : 
Luke Ἢ ὃ . : 

Paul’; . . . 
Hebrews c Ε - . - 


8 15. 
8 16. 
§ 17. 
§ 18. 
§ 19. 


1 Pet., James, Jude; 2 Pet. . 
Matthew, Mark 
The Johannine Writings . 5 . 


Apocalypse . : - δ “ . 


PAGE 


12 


12 
14 


14 
16 


REFERENCES. 


=— 


A FuLL list of authorities and references will be given with the con- 
cluding part of Vol. II. 

For collections of Papyri, Inscriptions and Ostraca reference must be 
made to the complete lists given in Moulton and Milligan, The Vocabulary 
of the Greek Testament (London, Hodder & Stoughton, 1914). 


ABBREVIATIONS. 


i 


In addition to those already explained in Vol. I. of this Grammar, 
the following may be mentioned :— 


Berl. Phil. Woch.= Berliner Philologische Wochenschrift. 
Blass—Debrunner= Friedrich Blass’ Grammatik des neutestamentlichen 
Griechisch. Vierte Aufl., von A. Debrunner. Gottingen, 1913. 
Blass Gr.= Grammar of NT Greek, by F. Blass. English tr. by H. St. J. 

Thackeray. Second edition, London, 1905. 

Blass Pron.= Pronunciation of Ancient Greek. English tr. by Parton, 
1890. 

Boisacg = Dictionnaire Etymologique de la Langue Grecque, par Emile 
Boisacq. Heidelberg, 1907. 

Brugmann‘ or Brugmann—Thumb=Grvechische Grammatik, von Karl 
Brugmann. Vierte vermehrte Aufl., von Albert Thumb. Miinchen, 
1913. 

CBE =Cambridge Biblical Essays, ed. H. B. Swete. London, 1909. 

CGT=Cambridge Greek Testament. 

DAC= Hastings’ Dictionary of the Apostolic Church. 2 vols., 1915, 1918. 

DCG = Hastings’ Dictionary of Christ and the Gospels. 2 vols., 1906, 1908. 

Debrunner—see Blass—Debrunner. 

Deissmann LAE=Light from the Ancient East. English tr. by L. R. M. 
Strachan. London, 1910. 

Deissmann Paul=St. Paul, A Study in Soctal and Religious History. 
English tr. by L. R. M. Strachan. London, 1912. 

xii 


XIV ABBREVIATIONS. 


Harnack Spriiche=Spriiche wnd Reden Jesu, A. Harnack. Leipzig, 1907. 

Harnack Sayings= The Sayings of Jesus, by A. Harnack. English tr. by 
J. R. Wilkinson. London, 1908. 

Harris Codex Bezae.=A Study of Codex Bezae, by J. Rendel Harris. 
Cambridge, 1891. 

Helbing Gr.=Grammatik der Septuaginta: Laut- und Wortlehre, R. 
Helbing. Gottingen, 1907. 

Kretschmer Hinl. = Die Einleitung in die Geschichte der griechischen Sprache, 
1906. 

Kretschmer Hntstehung= Die Entstehung der Kowwn (Sitzberichte ἃ. Wien. 
Akad., 1900). 

Kretschmer Vaseninschrifien=Dte griech. Vaseninschriften ihrer Sprache 
nach untersucht, 1894. 

Lake Cod. Sin.=Codex Sinaiticus Petropolitanus. With introduction by 
Kirsopp Lake. Oxford, 1911. 

Lipsius Gr. Unt.=Grammatische Untersuchungen tiber die bibl. Griicitét, 
1863. 

Lob. Paral. = Paralipomena Grammaticae Graecae, C. A. Lobeck. Leipzig, 
1837. 

Lob. Phryn.=Phrynichi Ecloga, ed. C. A. Lobeck. Leipzig, 1820. 

Meyer Gr.= Griechische Grammatik, von Gustay Meyer. Dritte vermehrte 
Aufl. Leipzig, 1896. 

Milligan Selections= Selections from the Greek Papyri, by George Milligan. 
Cambridge, 1910. 

Moeris= Moeridis Lexicon Atticum, ed. J. Pierson. Leiden, 1759. 

Moffatt = New Translation of New Testament. Third edition, London, 1914. 

Moffatt Introd.=Introduction to Literature of NT. Second edition, 
Edinburgh, 1912. (Third edition, revised, since published 1918.) 

Moulton Prol. or Proleg.2=A Grammar of NT Greek. Vol. 1., Proleg- 
omena. Third edition. Edinburgh, 1908. 

Moulton Hinl.=Einleitung in die Sprache des Neuen Testaments. (Trans- 
lated with additions from the third edition of Prolegomena.) 
Heidelberg, 1911. 

Moulton Christian Religion=The Christian Religion in the Study and the 
Street, by J. H. Moulton. London, 1918. 

Nachmanson Beitr.=Beitriige zur Kenntnis der altgriechischen Volks- 
sprache. Upsala, 1910. 

Oxford Studies= Oxford Studies in the Synoptic Problem, edited by W. 
Sanday. Oxford, 1911. 

Radermacher Gr.= Neutestamentliche Grammatik (Handbuch zum Neuen 
Testament, i. 1.) Tiibingen, 1911. 

Robertson Gr.= Grammar of the Greek Testament in the Light of Historical 
Research, by A. T. Robertson. London and New York, 1914. 

Rudberg Ntlicher Text=Neutestamentlicher Text wnd Nomina Sacra. 
Upsala, 1915. 

Sanders= Facsimile of the Washington Manuseript of the Four Gospels in 
the Freer Collection, with an Introduction by H. A. Sanders. Michi- 
gan, 1912. 


ABBREVIATIONS. XV 


Sophocles Lea.=Greek Lexicon of the Roman and Byzantine Periods, by 
E. A. Sophocles. Boston, 1870. 

Thackeray Gr.=A Grammar of the OT in Greek, 1., by H. St. J. Thackeray. 
Cambridge, 1909. 

Thumb Handb.= Handbook of the Modern Greek Vernacular, by Albert 
Thumb. Translated from the second German edition by S. Angus. 
Edinburgh, 1912. 

Thumb Dial.= Handbuch der griechischen Dvialekte, by A. Thumb. 
Heidelberg, 1909. 

Vocab.= The Vocabulary of the Greek Testament, by James Hope Moulton 
and George Milligan. London. Parti., 1914; part ii., 1915. 

Wackernagel Hellenistica = Hellenistica (Kinladung zur akadem. Preister- 
kiindigung). Gottingen, 1907. 

Walde Lat. Etym. Wort.=Lateinisches etymologisches Wérterbuch, von A. 
Walde. Heidelberg, 1906. 2 Aufl., 1910. 

Wesseley Zauberpap=C. Wesseley: Griechische Zauberpapyrt von Paris 
und London, in Denkschriften Wiener Akademie, 1888,°36, pp. 27—208. 

Witkowski=Lpistulae Privatae Graecae.*, ed. S. Witkowski. Leipzig, 
1911. 


DIT PVM a e) 


ope Fes 
+ 


ΓΝ 


A GRAMMAR OF NEW TESTAMENT GREEK. 
VOLUME IL. 


—— -- 


INTRODUCTION. 


§ 1. A Book which has already offered Prolegomena at con- 
siderable length and furnished with plenty of detail does not 
seem to need an Introduction when the systematic presenta- 
tion of grammatical material is at last about to begin. But 
there are very cogent reasons for procedure which may fairly 
enough be charged with cumbrousness. To begin with, my 
Prolegomena appeared in 1906, and much has to be added 
from the accumulations of a decade. The papyri and other 
sources have provided abundance of fresh material from which 
I could now enlarge the book much beyond the scale of 
the latest English or German edition. We have now the ad- 
vantage of discussion upon the views of New Testament Greek 
grammar which Deissmann’s pioneer studies in the vocabu- 
lary prompted. Without repeating what has been examined 
at length in the first volume, I may now apply the results to 
subjects which must be placed in their right light before we 
ean fill up the outline of Hellenistic grammar as it appears 
in the New Testament. I shall not tarry to repeat from Winer 
the history of earlier research in the subject: there is enough 
to do in delineating the conditions as we read them to-day. 


1. New TESTAMENT GREEK AS A UNITY. 


§ 2. That NT Greek is in general the colloquial lingua 
franca of the early Roman Empire has been made clear by 
the facts presented already, and we need not even summarise 


1 Many of the subjects discussed in this Introduction were sketched in the 
paper on ‘‘ NT Greek in the light of modern discovery,” in Cambridge Biblical 
Essays (ed. Swete, 1909). 

VOL, II.—I 


2 A GRAMMAR OF NEW TESTAMENT GREEK. [§ 2-3 


the case. With all the difference that there is between the 
writers of the NT, we can say of them collectively that they 
stand apart from literary Hellenistic monuments, the LXX 
excepted, in eschewing vocabulary, grammar and style which 
belonged to the artificial dialect of books, and applying to 
literary use the spoken Greek of the day. Their differences 
are comparable with those we notice between English speakers 
of varying degrees of education. Except for literal, and to 
some extent conventional translations, the NT contains no 
element which would strike contemporary Greeks as the 
archaic English of AV or RV strikes us to-day. 


§ 3. The first impulse to this use of the lingua cottidiana 
comes from the LXX. The Pentateuch, earliest and most 
important section of the Greek OT, quoted in NT so 
frequently as to show us at once how commanding was its 
influence, consists generally of good and easy vernacular 
Greek. In the day when it was made the tendency to 
Atticism had hardly begun to taint Greek literature. Literary 
Hellenistic was not colloquial in style, but it was no artificial 
dialect. Despite Aristeas, the LXX was not produced for 
learned consumption. The Greek OT, like the NT, was 
meant from the first to be the people’s book. When, there- 
fore, evangelists began to write down their story, or Christian 
preachers to compose informal pastorals for their Churches 
when far away, there was a precedent ready for their use of 
the popular speech. It was vital that they should write in 
language which would enable them to reach the widest 
audience at once. They could have used the literary dialect, 
some of them, at any rate. But Paul used the tongue of the 
unlearned for the same reason as John Wesley did: simple 
language is very easy for men whose one desire is to be clear 
and get their message home. Two centuries later Clement of 
Alexandria was Atticising for the same motive that made 
Paul Hellenise. Cultured people then would not read a 
book written in the vulgar tongue, and Clement was eager 
“by all means to gain some.” In the same spirit the apostles 
wrote as they spoke, that all might hear and understand. 
Their Greek represents, from the literary historian’s point of 
view, the greatest of those revolts against artificialism which 


§ 3-4] INTRODUCTION. 3 


have recurred through the ages and kept true literature alive. 
Just because Attic was the finest instrument human thought 
has ever played upon, the Epigoni tried to honour it by 
destroying the reality that gave it tone. The living daughter- 
speech was with them, tuneful and rich in all resources of 
expressiveness, though the foreign strain in her parentage had 
brought in some new intonations and lessened the delicate 
refinement of the mother-tongue. The taste of an age that 
could not understand refused to listen to the fresh young 
voice, and preferred to grind out ancient records on a gram- 
mophone. The Greeks are doing it still to-day, garnishing 
the mummy of the past instead of cultivating the rich 
resources of the present. Against this and every other such 
outrage on the spirit of literature the New Testament makes 
its protest. Only nature can give the touch which stamps 
the highest literature, and every book of the New Testament 
bears this mark beyond cavil. The Apocalypse is perhaps 
the extreme case. Its grammar is perpetually stumbling, its 
idiom is that of a foreign language, its whole style that of a 
writer who neither knows nor cares for literary form. But 
just because the weird dialect is the native speech of its 
author, if he must use Greek, we accept it without apology ; 
and no anthology of the rarest gems in human literature 
could be complete without contributions from its pages. 


§ 4. We shall have to differentiate presently between 
writers of very unlike culture and style, but a few summary 
words must be ventured as to the fitness of the Hellenistic 
vernacular as a medium for expressing what evangelists and 
apostles had to say. How does it compare with the languages 
which lie nearest, by nature or by circumstance? Take first: 
Semitic, in a dialect of which the NT might so easily have 
been written, since all its authors (except probably Luke 
and the author of Hebrews) counted Aramaic as their mother- 
tongue. The narrative parts, and such a book as the 
Apocalypse, would have suffered little. Lost Aramaic 
originals lie behind a fair proportion of these documents ; 
and if these treasures had survived, those familiar with the 
language might well have found them no less simple, forceful 
and vivid than the Greek which has supplanted them. It is 


4 A GRAMMAR OF NEW TESTAMENT GREEK. [S$ 4 


in the hortatory and doctrinal parts that the special ad- 
vantage of Greek appears. Equally capable of simplicity, it 
is capable of subtlety and precision beyond any Semitic dialect, 
and has a far wider range. We cannot imagine the founda- 
tion documents of Christian doctrine expressed in Old 
Testament Hebrew. Comparing Hellenistic with classical 
Greek, we may fairly say that the greater simplicity of the 
former gives it a decided advantage over even Attic for pure 
narrative, although the Ionic of Herodotus may claim equality. 
And it is fair to assert that what the Κοινή has lost of 
subtlety and grace, as compared with the Attic of the golden 
age, has been of little moment for the uses of the Christian 
writers. These elements are comparable with the more 
elaborate vocabulary which we find so highly cultured a man 
as Paul deliberately avoiding, as over the heads of simple 
people whom he wanted to reach. The characteristic 
strength of Greek was unimpaired—its wealth of significant 
differentiation in verbal tense system, its simple but adequate 
cases, made clear by prepositional resources which are no 
longer over-complex as in the earlier language. 

I might repeat here some words written in Cambridge 
Biblical Essays (500 f.) upon one significant instance :— 


The delicate precision of the use of the optative commands 
our admiration as we see it in the great writers of Athens. 
And yet we may remember that, except to express a wish, 
the optative has really no function which other moods cannot 
express equally well, so that by practically dropping the rest 
of its uses, Hellenistic has lost no real necessity of language. 
Indeed the fact that all the Indo-European dialects have 
either fused these two moods into one (as Latin) or let one 
of them go (as post-Vedic-Sanskrit), is evidence enough that 
classical Greek was preserving a mere superfluity, developing 
the same after its manner into a thing of beauty which 
added to the resources of the most delicate and graceful 
idiom the world has ever seen. But we are not belittling 
the masterpieces of Hellas when we say that their language 
was far less fitted than Hellenistic for the work that awaited 
the missionaries of the new world-faith. The delicacies of 
Attic would have been thrown away on the barbarians whom 


§ 4-5] INTRODUCTION. 5 


Paul did not disdain to seek for the Kingdom of Christ. If 
much of the old grace was gone, the strength and suppleness, 
the lucidity and expressiveness of that matchless tongue were 
there in undimmed perfection. They are recognised still 
when travellers master the unschooled “jargon” of the 
peasants in modern Hellas, the direct descendant of the 
᾿ Greek of Mark and Paul. As one of the most accomplished 
of them, Dr. W. H. D. Rouse, well says: “The most abstruse 
and abstract ideas are capable of clear expression in the 
popular speech. The book-learned will often hesitate for an 
expression, the peasant never. He spends all his days in 
talking, and has plenty of practice; and his vernacular is 
not only vivid and racy, it is capable of expressing any 
thought. . . . His language has the further advantage of 
being able to form new words by composition.” Assuredly a 
language which had all these characteristics three thousand 
years ago, and has them to-day, is scarcely likely to have 
lost them awhile during the great period when Greek was 
spoken and understood by a far larger proportion of civilised 
mankind than it had ever been in the period of its greatest 
glory, or has ever been again since East and West parted 
asunder and let the dark ages in. 


2. CONTACTS WITH LITERARY LANGUAGE. 


§ 5. The general rule that NT writers do not make use 
of the artificial literary dialect has one partial exception to 
prove it; and there are naturally degrees of approximation 
towards this dialect according to the extent of the writer’s 
education. We may take the exception first. It is a book 
which stands apart in many ways, by general consent 
decidedly the latest in the Canon, and the solitary NT 
example of pseudepigraphic writing. 2 Peter is written in 
Greek which seems to have been learnt mainly from books. 
Greek proverbs, Greek inscriptions,? and Greek books which 
we can no longer handle, contributed to the writer’s vocabu- 
lary, and moulded the fine sense of rhythm to which Mayor 
bears effective testimony. It is to literature rather than 
to vernacular inscriptions and papyri that we go when we 


1 See J. B. Mayor on 2”, 2 Deissmann, Bible Studies, 360 ff. 


6 A GRAMMAR OF NEW TESTAMENT GREEK. [8 5-¢ 


seek to illustrate rare words in this little book; and the 
general style is far removed from the language of daily life, as 
any tiro can see. These traces of elaboration are as much in 
keeping with the character of the book as the well-understood 
convention by which the writer shelters under a great name 
from the past. Only a shallow judgement could find in 
either the justification of disparaging views as to the Epistle’s 
value. The presence of a fair crop of solecisms is natural 
in a book so composed. If it was written, as generally 
supposed, early in the second century, we may note that the 
development of a language proper to books had advanced 
greatly since the age whence most of the NT writings come. 
As has been remarked already with reference to Clement of 
Alexandria, at the close of the same century,! the motive of 
this artificial language may well be that at the time of writing 
it commended a book to readers whose taste was no longer 
satisfied with a simple and natural style.” 


§ 6. It is a long step from 2 Peter to the Lucan writings, 
but we take them next because they and Hebrews alone 
show any consciousness of style. Hebrews, indeed, may be 
summarily dealt with as a composition into which admittedly 3 
nothing artificial has entered, though the writer’s culture 
prompts a style decidedly removed from the colloquial. 


1 For Clement’s Atticising see the monograph on his use of the optative by 
Jakob Scham, and my review in Deutsche Literaturzeitung, 1914, 1503-6. 

2 On Atticism as a literary phenomenon, reaching its climax in the second 
century A.D. and almost justifying itself in the hands of the brilliant Syrian 
Lucian, see especially Schmid’s Afticismus (Stuttgart, 1887-96). Its theory is 
seen best in Phrynichus (fl. A.D. 180), with his fine scorn of ἀμαθεῖς who (for 
example) could use γλωσσόκομον instead of γλωττοκομεῖον, and applied it to a 
box for books or clothes instead of restricting it to the mouthpieces of flutes. 
W. G. Rutherford’s New Phrynichus (London, 1881) edits the old pedant for 
us, and adds thereto many like words. 

3 Blass, indeed (Brief an die Hebrder, Text mit Angabe der Rhythmen, 
Gottingen, 1903; cf. Grammatik der Neutestamentlichen Griechisch”, 304 f.), 
argued for an elaborate system of rhythm in Hebrews, which would have trans- 
ferred the Epistle into the literary category very decisively. It may be very 
seriously questioned whether prose rhythm was consciously elaborated even by 
the Attic orators, from the study of whom Blass derived his theories: it is 
probable that instinct alone trained the ear to rhythm, even when analysis can 
formulate rules. That Blass could discover orthodox rhythms even in Paul might 
fairly count as a reductio ad absurdum of his theory for Hebrews. 


§ 6] INTRODUCTION. 7, 


The absence of the potential optative is a primary test of 
freedom from artificialism, and this is complete in the Epistle. 
The best analogue will be the pulpit style of a cultured 
extempore preacher, or that of his letters to the religious 
press. The test just mentioned needs further inspection for 
Luke, the only NT writer to use the potential optative, in 
indirect questions and conditional with av. The latter is still 
used in the epistolary formule of Ptolemaic times, when the 
writers are well educated, and it can hardly be called 
artificial, though in i/A.D. it must have been almost confined 
to book language. Since the growth of the Atticising 
movement was bringing the optative into greater prominence 
as a literary usage, it may be safely said that the presence of 
this survival was by this time essential for any claim to 
style. We are left then with Luke as the only /ittérateur 
among the authors of NT books. (I make no apology for 
speaking of “Luke”: those who prefer “Lk,,” “Lk,” .. . 
“ Lk,,,” are, of course, welcome to their opinion. I would only 
observe that in grammar and vocabulary and phraseology 
Lk, 2 . τ . » have anastonishing resemblance to one another.) 
In using the term we are not suggesting that Luke capitulated 
to the growing fashion of going back to archaic models as 
alone suitable for literary composition. A page of Josephus 
would disabuse our minds of any such idea. It is only that 
Luke as a Greek fell by a native instinct into the habit of 
style which would make his narrative tell. It would be 
hard to find ancient parallels for the variation of style he 
shows as his story changes its scene. A modern novelist 
will see to it that his country yokel and his professor do not 
talk the same dialect; and he will often try to make a 
Lancashire weaver or a Cornish miner approximate to the 
speech actually current in those areas. Similarly, Aristo- 
phanes makes a Megarian, a Beotian, a Spartan woman 
speak their own dialect fairly correctly. But this is only 
partial illustration: it suits Luke’s accurate reproduction of 
the reported dialogues that came to him in rough translations 
like that we postulate for Q@. But it is not going as far as 
Luke when he steeps his style in Biblical phraseology, drawn 
from the Greek Old Testament, so long as his narrative 
moves in Palestinian circles, where the speakers use Greck 


8 A GRAMMAR OF NEW TESTAMENT GREEK. [8 6-7 


that obviously represents a foreign idiom—like Shakespeare’s 
Fluellen with his Welsh English. That Luke should do this 
fits in well with his presumed history. A proselyte who 
made his first acquaintance with the Old Testament in its 
Greek version was likely to feel for that version as no 
Hebrew could feel, accustomed to keep all his reverence for 
the original. His imitation of the translation-Greek of his 
model—eg. in the construction καὶ ἐγένετο καί with a finite 
verb, which yields to the ace. et infin. in Ac '—reminds us of 
the Biblical style of John Bunyan, and other English writers 
whose education it was.to be homo unius libri. That Luke 
instinctively departs from that style when his subject takes 
him away from the Biblical land and people, is equally 
natural. It is mostly in these parts of his work that he 
makes what concessions he does make to the book style. 
We are sometimes able to distinguish between the Greek of 
his sources. Compare the masterpiece of Lk 15 with the 
parable that follows. There is absolutely nothing in the 
story of the Two Sons which suggests translation from a 
Semitic original: the conjecture rises to one’s thought that it 
never was translated, but spoken in Greek to an audience 
that knew no Aramaic—a point to which we shall return. 


§ 7. There is only one other writer whom we might 
expect to show contacts with the literary Greek. A highly 
educated man like Paul,? who spent his early years in a great 
centre of Hellenistic culture, might have used the book Greek 
as to the manner born. It is very obvious that he did not. 
The exordium of an address to Athenian philosophers 
survives to show us that he could use the language of the 
higher culture when occasion required? But his letters, 


1 See Proleg., 16 f. 

2 Professor Deissmann’s brilliant work, St. Paw/, seems to miss the mark 
altogether in describing Paul as a working man, largely on the strength of his 
big clumsy writing inferred from Gal 6". If this interpretation of πηλέκοις 
γράμμασιν be conceded, such writing does not nowadays imply illiteracy, and 
we have no evidence that it did in Paul’s day. The Apostle’s tent-making is 
completely explained by a well-known precept of the Rabbis, and his exercise 
of the art by Ramsay’s most reasonable supposition that a bigoted Jewish 
father had cut him off. 

3 Of course Luke is usually credited with Paul’s Areopagitica, and it may be 
difficult to prove completely that he wrote his report from full notes, given 


§ 7] INTRODUCTION. 9 


addressed to churches into which “not many wise were 
called,” are studiously kept within the range of popular 
vocabularly and colloquial grammar. Nageli’s monograph ! 
shows this conclusively for the vocabulary. As to the 
grammar, it may be noted that Paul uses the highly 
colloquial types yéyovay, and perhaps παρελάβοσαν 5 (the 
former also Luke), also καυχᾶσαι, ζηλοῦτε and φυσιοῦσθε as 
subj., νοΐ, ἐφ᾽ ἑλπίδι, ete. In the use of popular forms he 
and Luke go as far, with rare exceptions, as the least cultured 
of NT writers. These facts are the strongest possibie 
disproof for both Paul and Luke of any charge of using book 
Greek: no author who could favour the Atticist rules would 
fail to purge his pages of vernacular inflexions. If, however, 
the two friends keep company in their inflexions, they part 
again in vocabulary and in so typical a matter as the use of 
the optative, and in both Paul leans away from the literary 
style. What Paul might have done had he been writing 
“ treatises ” (Acyoc—Ac 1!) like Luke, we cannot say. What 
has come down to us from him is all of a casual character, 
open letters to communities, for which permanence was 
never contemplated. There is a good modern parallel in 
Wesley’s Sermons, addressed to plain folks in simple 
language: even when the Fellow of Lincoln preached before 
his University, he took care, in republishing the sermons in a 
volume destined to be an informal manual of doctrine, to 
keep their language within popular range. In them as in 
the Pauline Epistles “ignorant and unstable men” might 
find δυσνόητά τινα, but it would not be because of their 
dialect. 


him not long after by his master. But when we find the Lukan Paul quoting 
Epimenides (Ac 17**), and the Paul of the Pastorals citing the very same 
context (Tit 113), with the Aratus-Cleanthes quotation (7b."5") to match the 
Menander (1 Co 1638), we may at least remark that the speech is very subtly 
concocted. Paul was, moreover, much more likely than Luke to know the 
tenets of Stoics and Epicureans so as to make such delicately suited allusions 
to them. Luke’s knowledge of Greek literature does not seem to have gone 
far beyond the medical writers who so profoundly influenced his diction. He 
no doubt shared with all educated Greeks some familiarity with Homer: the 
obsolete word ναῦς in Ac 274! was acutely traced to Homer by Blass (Philology 
of the Gospels, 186). 

1 Das Wortschatz des Apostels Paulus (Gottingen, 1905). 

2 Proleg. 52: I have modified my view with the accumulation of evidence. 


10 A GRAMMAR OF NEW TESTAMENT GREEK. [88 


§ 8. The remarkably good Greek of James and First Peter 
will engage our attention under another heading, but we may 
add here one or two points which suggest themselves in 
another Palestinian writer, the author of the First Gospel. 
In spite of Harnack, there seems little doubt that he alters 
the language of his sources very much more than Luke does, 
so as to make the style of his work decidedly more uniform. 
He shows the artist in his genius for compression,! and in 
his fondness for Hebraic parallelism ;? while he frequently 
substitutes literary flexions for popular. Thus where Q ap. 
Luke has ἀμφιάξει, Mt 6°° shows the obsolete ἀμφιέννυσιν, 
where it has συνάξαι (Lk 3!” &*), Mt 3” dexterously brings 
in the future συνάξει, as in ἐπισυνάξαι, Lk 1385, mended to 
ἐπισυναγαγεῖν in Mt 23%", 

The degree of literary flavour attained in all these 
amounts to very little. It may be compared, on the one 
side, to our literary avoidance of colloquialisms like don’t and 
can't, which everybody uses almost exclusively in conversa- 
tion, but instinctively replaces by the /ento forms in written 
style, except in private letters. On the other side, we have 
a whole vocabulary which has its perfectly natural place in 
written English, with the same exception, and in the higher 
spoken style of serious oratory, but strikes us instantly as 
pedantic or affected when brought into conversation: Greek 
words of a similar type are avoided by Paul, but used by 
Luke and the auctor ad Hebreos. The mere fact that no 
NT writer thinks of avoiding the flexions which conspicuously 
distinguish Hellenistic from Attic,t or of using the dual, the 
final optative, or other Atticist hall-marks, is enough by itself 
to show that even though NT writers might sometimes take 
some pains with their style, the better to achieve their purpose, 
they would never allow themselves an archaism or affectation 


1 Cf. 3" where βαστάσαι, take off, expresses the full content of four words in 
Mk 17 and 11” where ἐπιγινώσκει exactly represents γινώσκει τίς ἐστιν of Q, etc. 

*See my paper in Kapositor, VII. ii. 97 f. (reprinted in J. H. Moulton, 
The Christian Religion in the Study and the Street, 47 ff. ; ef. ἐδ. 79). 

3 See Camb. Bibl. Essays, 485. 

4Such as the types σπείρης, νοός vot, ἡμίσους, ἀπεκατεστάθη, ἐλελύκεισαν, 
γένοναν, dot, ὀδυνᾶσαι, φάγεσαι, λυέτωσαν, oldas οἴδαμεν, ἤμην, ἐλήμῴθην, ete. 
There are, of course, some which only Mark or the Apocalyptist would admit, 
such as λέλυκες or ἔλυσες. 


§ 9] INTRODUCTION. 11 


which might endanger their being “understanded of the 
people.” 


§ 9. One further point needs to be guarded. There are 
some tests of literary Greek which have been applied in 
misapprehension of the facts and have produced results that 
are wholly misleading. Such is especially the assumption— 
treated as axiomatic by Harnack—that compound verbs are 
an evidence of cultured Greek. Harnack! builds upon it one 
of his working principles in reconstructing Q out of Mt and 
Lk: where either of them (which usually means Mt) shows 
the simplex, it goes back to Q, which is assumed to be written 
in rude vernacular. The axiom fails to survive so elementary 
a test as the counting of compound verbs in Mk and Lk. It 
is found that the two evangelists have an identical percentage 
per page, while their Greek notoriously differs more widely 
than anything else within the limits of the NT. Mark has 
actually 5°7 compound verbs per page (of WH), while Acts 
has 6°25, Hebrews 8°0, and Paul only 3°8. Reference may 
be made to the statistical investigation in Camb. Bibl. Essays, 
492 f., where it is shown that illiterate private letters among 
the papyri employ compound verbs as conspicuously as Mark. 
A fondness for compounds is fairly enough noted as a 
characteristic of an individual style: for example, the 
contrast between the figures for Paul and for Hebrews (Ὁ 8 
per page and 8.0) is enough to discredit the Pauline author- 
ship of the Epistle, were there further need of witnesses. But 
Harnack’s test must clearly disappear from our critical 
tool-box, The real history of the matter is that the increased 
use of compounds was one of the features of the Κοινή as 
compared with classical Greek,? and applied to literary and 
vernacular language alike. Writers like the First? and 
Fourth Evangelists, who markedly prefer simplicia, are in 
this regard aloof from a prevailing tendency. 


1 Sayings, 150 ; see the German Spriiche, 106. 

2 Compare the fact that Balyw simplex is very nearly extinct in Hellenistic : 
see Vocabulary, s.v. Mark uses compounds of πορεύομαι, the simplex 
never. 

8 Note as a typical example Mark’s ἀνεμνήσθη in 147, which Mt 267 reduces 
to ἐμνήσθη, while Lk 22° varies it to ὑπεμνήσθη. 


12 A GRAMMAR OF NEW TESTAMENT GREEK. [8 10 


3. SEMITIC COLOURING. 
(1) Language Conditions of Palestine. 


§ 10. There is a large bibliography on the problem of 
“the mother-tongue of Jesus.” We cannot discuss the 
problem here, but summary statements of results are 
demanded. It seems to me highly probable that some of 
the contradictory data may be reconciled by making more of 
the difference between Jerusalem and Galilee. The Holy 
City was in our Lord’s time a metropolis of aggressive 
nationalism. The Hellenising high priests, who had filled 
the city with Greek customs and speech, were no more; and 
a fanatical hatred of all things foreign was limited only by 
the hard fact of a Roman Procurator and soldiers at his 
command, within striking distance of the Temple. | Under 
such conditions it is easy to see that a knowledge of Greek 
would be reduced to a minimum demanded by the necessities 
of intercourse with pilgrims from the Dispersion and officials 
of the Roman government. Galilee, on the other hand, was 
notoriously “of the Gentiles.” There were towns there, 
such as Tiberias, where Jews and Judaism were invisible. 
Two centuries before, there had been a general clearance of 
Jews, and the consequences were sufficiently lasting to give 
a cue to modern paradoxists like Herr Houston Stewart 
Chamberlain, who would fain convince us that as a Galilean 
Jesus must have been of Aryan blood. The swine of Gerasa 
(Mk 5"), when all is said, do not approve themselves as 
belonging to Jewish masters. The question really is what 
language or languages did the Gentile majority in Galilee 
speak in the first century, which the Jewish minority were 
compelled to use if they had any dealings with them. Now 
Aramaic was not only the special language of the Jews: it 
was in rapidly growing use as a lingua franca in Western 
Asia, its rival, of course, being Greek. The realm of Greek 
as a world-language extended far beyond Palestine at the 
time of its greatest influence. Greek inscriptions are found 
all over Asia Minor and eastward up to the borders of India 


1 Foundations of the Nineteenth Century, i. 210; cf. Paul Haupt, Z’ransac- 
tions of the Third International Congress for the History of Religions, i. 304. 


§ 10] INTRODUCTION. 12 


—as far, in fact, as Alexander’s arms had penetrated.1 Two 
parchments have been published lately? which came from 
Avroman in Media, dated respectively B.c. 88 and 22. 1. 
They are the title-deeds of a vineyard, and are written in 
good Κοινή Greek: with them, however, is a third in a 
dialect unknown, the document being still undeciphered, 
In his notes Mr. Minns ealls attention to the fact that 
the tide of Greek language supremacy began to ebb from 
Western Asia about the beginning of the Christian era. 
Aramaic or Syriac would be the natural tongue of 
Gentiles as well as Jews in Galilee a very few generations 
later. But it does not appear that Greek was expelled, 
or near expulsion, in the early part of 1(Δ.Ὁ. We are at 
liberty then to reflect on the notable fact that three NT 
books are traditionally assigned to Galilean writers, viz. 
1 Pet, Jas and Jude, and that their Greek—especially that 
of the first two—is of a remarkably free and idiomatic kind. 
On the other hand, Mark was a Jerusalemite, and his Greek 
equipment is very meagre. The John of the Fourth Gospel 
and the Epistles is also on very strong grounds claimed as a 
Jerusalemite, and his Greek, while correct enough, is very 
bald and destitute of idiom. May we not infer that Galileans 
might be expected to use Greek freely, as having been 
accustomed to it by living among Greek-speaking people ? 
This does not mean that we should question the usual 
assumption that the bulk of our Lord’s teaching, public 
and private, was in Aramaic, the language from which 
Palestinian Jews were not likely to deviate except when 
speaking to people who only knew Greek. But that He 
and His disciples were thoroughly familiar with Greek seems 
altogether probable. It is evident that if Mark’s indifferent 
Greek may be credited to his Jerusalem upbringing, we 
have a contributory item which may be useful for some 
critical questions. 


1 See Dittenberger’s two volumes, Orientis Greci Inscriptiones Selecte 
(Leipzig, 1903, 1905). 

2 Ellis H. Minns, ‘‘ Parchment of the Parthian Period from Avroman in 
Kurdistan” (JHS, 1915, 22 ff.). 


14 A GRAMMAR OF NEW TESTAMENT GREEK. [811] 


(2) Aramaisms and Hebraisms. 


§ 11. The past decade bas produced much helpful dis- 
cussion on the burning question of Semitism in the Greek 
Bible. Nothing has emerged, I believe, to shake the general 
position taken by Deissmann, adopted with some develop- 
ments in Prolegomena, but there are some applications of the 
principle which I should myself admit to be too rigorous. 
It will be advisable therefore to restate the central thesis of 
“ Deissmannism,” albeit, alas! without the advantage of Pro- 
fessor Deissmann’s own judgement, to seek which was in happier 
days as much a pleasure as a duty. 

Semitism in the NT will be defined as a deviation from 
genuine Greek idiom due to too literal rendering of the lan- 
guage of a Semitic original. “Semitic” for this purpose 
means either Hebrew, as the language of the Old Testament, 
or Aramaic, as the mother-tongue of many NT writers. The 
definition omits intentionally the case in which literal render- 
ing of Semitic produces Greek which is perfectly idiomatic. 

The resulting sense may be (1) identical. In that case it 
might have seemed that we were spared the trouble of discuss- 
ing Semitism, unless we felt ourselves bound to find “ Latinism ” 
in the sentence “ Balbus built a wall,” which is an undeniably 
literal rendering of Balbus murum edificawt. E. Nestle, 
however, a first-rate authority on Semitic subjects, stoutly 
claimed ἕως πότε ; as a Hebraism, “even if it is still used by 
Pallis in his MGr translation,” and though it “may be 
quotable from early Greek, and have spread in later times.” 
To this declaration, put forth in a review of my Prolegomena, 
I replied with the question whether the Emperor Hadrian’s ἐκ 
πότε; and our own (till when? were likewise to be branded as 
Hebraism.! Of course, all languages when we compare them 
show multitudes of idioms in which two or more of them 
exactly agree. The generally similar structure of the human 
mind secures this mitigation of the translator’s otherwise 
intolerable lot. But beside this case, which really does not 
deserve detailed investigation, there is the more difficult case of 
approximation not amounting to identity. A literal or nearly 


1 Camb. Bibl. Essays, 473 f.; Proleg.® 107. 


§ 11) INTRODUCTION. 15 


literal rendering may give us a phrase which is moderately 
idiomatic, but of decidedly restricted use in the language 
of the translation. The result may be a very marked over- 
use of a rare locution, as representing exactly what is common 
in the language from which the translation is made. Thus 
the very rare preposition é€ywmvoyv—the adverbial neuter of 
an adjective found in Greek literature, though seldom enough 
—figures in legal Greek papyri to represent coram, without a 
case expressed, and in some NT writers to render "259, It is 
quite genuine Greek, but it is fair to call it a Latinism in the 
papyri and a Hebraism in Luke, since it is most unlikely that 
either would have used it except in reference to its original. 
So again the relative frequency of ἐδού in James—compared 
(Proleg. 11 n.) with the Welshman’s “look you ” in Shakespeare 
—may be reasonably enough called Hebraism if we only mean 
that its prominence is to be connected with the writer’s familiar- 
ity with a language in which an interjection with this mean- 
ing was used much more frequently than it was in native 
Greek. While, however, we are justified in considering all 
such cases of “ over-use”” when we are estimating the language 
of a particular writer, it would be well to restrict the term 
Semitism (Hebraism, Aramaism) to cases where Greek idiom 
is violated or at least seriously strained. We will add the 
adjective “secondary” when Semitisms of the milder kind 
are in question. 

Then (2) the resultant meaning, when literal translation 
has produced idiomatic Greek, may be something different 
from that of the original. In this case, of course, the trans- 
lator must have misunderstood his original, or else failed to 
realise in what sense ordinary Greek readers would under- 
stand his phrase. A good example of the former was pointed 
out by Thackeray! in Lk 14”. It is clear that when we 
meet in a NT book a phrase which makes good sense as Greek, 
we shall have to treat it as Greek: we may sometimes 
suspect that the writer was really thinking of something 
different, and we may have evidence from his lapses elsewhere 


1 JTS xiv. 389f. Here the reading of B preserves the recognised translation- 
Greek of the later LXX for the familiar Hebrew phrase for salutation, which 
where royalty was concerned acquired the special connotation of tendering 
allegiance, or, as in this passage, of making unconditional surrender. 


16 A GRAMMAR OF NEW TESTAMENT GREEK. [S 11-12 


which makes the suspicions plausible, but obviously the Greek 
readers for whom the book was intended never suspected 
anything of the kind. Except in the Apocalypse, where we 
have a writer who simply did not lmow the grammar of 
Greek except in shreds and patches, we shall hardly care to 
allow that the readers of the book on its first appearance had 
no adequate equipment for understanding what the author 
meant; and even in that book we shall only admit the 
assumption very sparingly. We may take as an example 
Wellhausen’s treatment of Mk 27:1 λαλεῖ βλασφημεῖ are to 
be taken together as a blundering attempt to represent an 
Aramaic construction which would be accurately rendered by 
λαλεῖ βλασφημίας (Lk 5%).2 Whether this is the most prob- 
able Aramiac original we need not inquire: it is enough to 
reply that no Greek reader could possibly suspect any other 
sense than that which the RV represents, and that Luke’s 
paraphrase is no warrant for making Mark guilty of a wildly 
impossible Greek combination, with no second offence to create 
a presumption against him. The fascinating pursuit of 
Aramaic originals may lead to a good percentage of successful 
guesses; but they are mere guesses still, except when a 
decided failure in the Greek can be cleared up by an Aramaic 
which explains the error, and this acts as corroboration. 


§ 12. True Semitisms in the NT are of two kinds. First 
come imitations, conscious or unconscious, of the Greek OT, 
where the translators had perpetrated “ translation Greek.” 
Secondly, there are similarly slavish renderings of Semitic 
sources, oral or written, which lie behind the NT documents: 
we may here stretch the term “sources” to include a writer’s 
native Semitic in which he frames his sentences in his own 
mind, and then more or less successfully translates them into 
Greek. Of course, in the OT the Semitisms only differ from 
the second class just named in that they are Hebraisms, 
while those in the NT are Aramaisms—Aramaic originals in 
OT and Hebrew possible originals in NT may be left out 
of account. Perhaps we should add the difference due to the 


1 See his Hind.} 22. 
2 Matthew characteristically abbreviates: οὗτος βλασφημεῖ (9°) practically 
contains the sense of τί οὗτος οὕτω λαλεῖ; βλασφημεῖ---866 above. 


§ 12] INTRODUCTION. Τὴ 


fact that the LXX is a definite translation of a series of books, 
long current and highly authoritative, while in the NT we 
have free composition in Greek, based frequently upon Semitic 
which had no fixed or authoritative form. The NT Aramaisms 
accordingly will be unconscious, and due to defective know- 
ledge of Greek. The Hebraisms of the LXX were very often 
conscious sins against Greek idiom, due to a theory that words 
believed to be divinely inspired must be rendered so that every 
detail had its equivalent. It was this which gave birth to 
Aquila’s ἐν κεφαλαίῳ ἔκτισεν ὁ θεὸς σὺν τὸν οὐρανὸν καὶ σὺν 
τὴν γῆν: no Greek could imagine what the σύν meant, but 
the Hebrew D8 must not be left without an equivalent. It 
must be admitted that our own RV was as unhappily con- 
scientious when it gave us “ By hearing ye shall hear,” or “ who 
also have been in Christ before me” (Rom 167). Translation 
of this kind is, of course, an outcome of conditions peculiar to 
canonical books. In the LXX we find very little of it in the 
Pentateuch, executed before this theory of a translator’s duty 
was framed, and very little in a book like Zobit, which only 
became (semi)canonical in its Greek, or rather in one of its 
two Greek forms. In estimating the effect of the LXX 
upon NT language we have to note carefully the very 
different degree in which its various parts influenced NT 
thought. 

If we count the separate verses cited in WH to make a 
rough test, we find that the Pentateuch accounts for a quarter 
of the New Testament quotations and allusions, the Prophets 
(and Daniel) for nearly a half, and the Psalms for a fifth, 
while all the rest only amount to 6 percent! It may be 
added that Jsaiah claims two-fifths of the proportion credited 
to the Prophets. Putting aside, therefore, the relatively 
negligible historical and poetical books, we have two forces 
acting on the NT writers from the Greek OT. On the one 
side is the good Kown Greek of the Law, the work of men 
who understood their original thoroughly, and aimed at 
expressing its meaning in plain every-day speech. On the 
other, there is the often inferior Greek of the Psalms and the 
Prophets, where the much more difficult original was frequently 


1 Camb. Bibl. Essays, 475. 
VOL. II.——2 


18 A GRAMMAR OF NEW TESTAMENT GREEK. [8 12-13 


misunderstood, and the misunderstanding often veiled by 
slavish literalness, while the development of the more rigorous 
theory of translation introduced yet more of this Greek that 
was no Greek. Since quotations from Prophets and Psalms 
are between two and three times as numerous as those from 
the Pentateuch, we might expect to find the stylistic influence 
of the latter altogether counterbalanced by the linguistically 
mischievous effects of the former. But the NT writers, 
except probably Luke and the author(ess) of Hebrews, knew 
the HeLrew original too well to be at the mercy of a defective 
translation. We very rarely find quotations which seriously 
violate Greek idiom. The “ Biblical style” which influenced 
pre-eminently Luke among NT writers was that of the Penta- 
teuch. It came first in time, stood first in authority, and 
being very largely narrative was more calculated to affect 
narrative books than the other books, which mostly supplied 
isolated phrases for quotation." 


§ 13. We proceed to remark on the extent to which 
Semitisms and secondary Semitisms may be observed in the 
several writers.” Let us take Luke first, both as the largest 
individual contributor and as the one who exhibits specimens 
of Hebraism to an appreciable extent. The most typical of 
Luke’s many imitations of OT Greek is the narrative “ it 
came to pass”—to represent it by the Biblical English, the 
appearance of which in one of our own writers would produce 
almost exactly the same mental association. How far this 
locution approximated to vernacular Greek idiom has been 
discussed in Proleg. 16 f.; and the significant fact is noted 
that in Acts Luke reverts to the form which least diverges 
from that vernacular. Luke often goes further in imitation 
of the Hebrew by writing καὶ éyéveto ἐν τῷ ὁ. infin. καὶ... 
here, also, there are cases elsewhere, as in Mk 44, where both 
Mt and Lk agree in omitting. Apart from these imitations 
of the Greek Bible, Luke shares with others certain 
Aramaisms which arise from literal rendering of vernacular 
sources. Whether Luke himself or his own immediate 


1 For some further remarks on LXX Greek, see my already cited essay in 
Camb. Bibl. Essays, 475 f. 
2 The details will be reserved for the Appendix. 


§ 13] INTRODUCTION. 19 


sources in Greek were. responsible can hardly perhaps be 
decided dogmatically. Neither Aramaic specialists nor 
Hellenistic have the right to decide whether he had any 
knowledge of a Semitic tongue: what we really need is pro- 
longed collaboration of both, till a joint impression is formed 
which may have elements of authoritativeness. Much de- 
pends upon our opinion as to Luke’s antecedents. If he was 
an Antiochene, he might very well speak Aramaic, as a 
language already beginning to dispute with Greek the position 
of general medium of communication all over Western Asia. 
If he was a Philippian, which seems to me very much more 
probable,? he would have to learn Aramaic in Palestine, 
which he seems to have visited first in 57 a.p. His.“ two 
years” (cf. Ac 2451) in the country were doubtless the oppor- 
tunity of collecting material for his Gospel and the earlier 
part of Acts. Did he trouble to acquire Aramaic for the 
purpose? It was in any case not essential, for in Galilee 
Greek-speaking people abounded, and even in Judwa—if 
Luke’s researches were really pursued there, of which there 
is not much evidence—it would be very easy to find inter- 
preters. If this is true, all Luke’s Palestinian material could 
come to him in Greek, and any Aramaisms or other phenomena 
traceable to defective Greek may be transferred to the various 
informants whose contributions Luke scrupulously noted down 
and reproduced. When, on the other hand, we find evidence 
that Luke’s text involves a misunderstanding of a Semitic 
original, such as would often occur when a foreigner with a 
fair but incomplete knowledge of the native dialect gathers 
information from people of varying degrees of education, it is 
obvious that such misunderstanding may as easily be credited 
to Luke’s sources as to himself. To prove him responsible, 
we should at least have to show that they were very 
numerous and evenly distributed, and that the same kind 
of mistake occurred in different places. And even then it 


1 See § 10. 

* The tradition of his connection with Antioch appears in Eusebius (HH 
iii. 4. 6), Jerome (de Vir. Iilustr. vii.), and the Monarchian Prologues 
(Kleine Texte i., by H. Leitzmann, Bonn, 1902). See art. ‘‘ Luke,” by K. Lake, 
in DAC. For the view that Luke belonged to Philippi, see Ramsay, St. Paul 
the Traveller, 200 ff. ; art. ‘‘ Luke,” by Souter, in DCG. 


20 A GRAMMAR OF NEW TESTAMENT GREEK. [8 18 


might only mean that Luke took about with him some 
Christian brother as his dragoman, a Greek who had been in 
the country longer than himself and had a passable know- 
ledge of Aramaic. 

That the two chief sources, used by Luke and by the 
First Evangelist, were Greek, is, of course, admitted. Mark’s 
defective Greek supplied Luke with Aramaisms ready made ; 
and sometimes a phrase of Mark’s, by which an Aramaic 
idiom is rendered word for word, the corresponding Greek 
idiom being inaccessible to him, may produce misunderstand- 
ing on Luke’s part. Equally assured is Luke’s use of a Greek 
Q, one of the translations of the Apostle Matthew’s Logia, as 
Papias’s famous sentence prompts us to hold.1 Here we are 
constantly finding that Luke faithfully preserved the rough 
Greek of his original, where Mt freely edits.2 That Luke 
treats his other sources along similar lines does not even 
depend on the acceptance of this doctrine, which I do 
not pretend to state as an admitted fact, though it seems 
to me quite certain. The wide differences in Greek style 
between one section and another of Luke’s peculiar matter 
can only be explained by assuming that he reproduced his 
sources generally as he received them. Probably this 
was mainly because in reporting discourses of Jesus he 
felt it was the safest procedure, since he had no materials 
for checking his sources. He had “revised them afresh” 
(18 παρηκολουθηκότι ἄνωθεν) with personal inquiry; but 
when he had no information enabling him to improve what 
he felt to be defective, it was better to copy his notes as 
they stood than to amend them by guesswork. It is the 
existence of these wide divergences between the discourses — 
in Luke’s peculiar sections which weighs most with me in 


1 We probably do not make enough of his very definite assertion that 
‘* every one translated the Logia as best he could.” While ‘‘ Matthew” and 
Luke certainly used the same version for a considerable number of the sayings, 
for which we are justified in using the common symbol Q, it is highly probable 
that they often had different versions, and that with this in mind we should 
spare our ingenuity superfluous exercise in places where Mt and Lk widely 
differ. 

2 For some argument in support of this thesis, against Harnack, I may refer 
to Expositor, VII. vii. 411 f. (or Moulton, Christian Keligion, 71 f.). One or two 
typical minutiz are repeated above, p. 10. 


§ 13-14] INTRODUCTION. 21 


my own judgement that Luke knew no Aramaic. Had he 
been his own translator, we should have expected to find 
the same evenness in the distribution of Aramaisms as we 
find in those general features of grammar and style which 
so overwhelmingly vindicate the unity of the two books ad 
Theophilum. 


§ 14. We pass on to Paul, the next largest contributor to 
the NT Canon. It is soon realised that we have no longer 
to do with effects of conscious style. Opinions may differ as 
to the proper description of Romans, his weightiest work : 
some insist upon its casual character as an open letter 
addressed to a church that needed doctrinal upbuilding, with 
no more elaboration than we should put into a letter to the 
press, while others would make it approximate to a set 
treatise. But even if the second alternative were adopted, 
there is no possibility of claiming any definitely literary form. 
Niaigeli’s study of a section of Paul’s vocabulary shows that he 
kept himself to words in popular use. Similarly in grammar 
and style we look in vain for constructions or inflexions of an 
archaic or worked-up character. As to his Greek, it is 
obvious from all we know of him that he must have spoken 
Greek from the first as freely as Aramaic. He calls himself 
"E8paios ἐξ ᾿Εβραίων, “a Hebrew of Hebrew descent,” and 
the term naturally implies the familiar use of the Semitic 
mother-tongue, But the most patriotic Jew of the Dispersion 
could not get on without Greek. It need not be added that 
for Paul’s missionary work in the West, Greek had no possible 
alternative except Latin. A man thus accustomed to use the 
language of the West was not likely to import into it words 
or constructions that would have a foreign sound. The LXX 
had no such supreme authority for Paul that a copying of its 
language would strike him as natural. And if Greek was an 
alternative mother-tongue to him, he would use it too un- 
consciously to drop into Aramaisms, defective renderings of a 
language he could correct as well as any one. The a priori 
view thus sketched tallies satisfactorily with the observed 
facts. Paul very rarely uses phrases which come from a 
literal rendering of the Semitic. His Semitisms are secondary 
at most—defensible as Greek, and natural to a Greek ear. 


22 A GRAMMAR OF NEW TESTAMENT GREEK. [814 


How carefully he kept away from language which might seem 
archaic or remote to the ordinary people for whom he wrote 
is well seen in the case of aéns.!_ Paul deliberately mars the 
rhetorical effect of the quotation from Hosea in 1 Co 15* by 
substituting θάνατε for ἅδη. For Hades was a Greek divinity, 
not a place, and the name, though common enough in litera- 
ture, had dropped out of the ordinary vernacular. Its 
occurrence elsewhere in the NT may be traced directly to the 
influence of the LX X, where it is freely used. The LXX 
translators appropriated it from the technical language of 
Greek religion because they found it to be an exact render- 
ing of the Hebrew ORW, For Paul, however, no such felicity 
in the commerce of tongues could stand against the plain fact 
that the word in question had no place in the vocabulary of 
every-day Greek. In turning from lexical to grammatical 
considerations of style we may look at one or two hall-marks 
of Semitism as they affect Paul.2 Twice in Ephesians we 
find the collocation πᾶς οὐ or μή for οὐδείς or μηδείς, which 
has been quoted as “a sign that the Semitic influence passed 
from Paul’s thought into his language.” But this Hebraism 
never occurs elsewhere in Paul, and its occurrence twice in 
this one disputed Epistle must at least be put among the 
special features of its language which have to be explained.® 
In 5° there is another possible Hebraism, ἔστε γινώσκοντες, 
the coincidence of which with was... ov in the same 
sentence perhaps emphasises the presence of language moulded 
on Biblical phraseology. Hphesians keeps to itself almost as 
completely another noteworthy Semitism—*“ sons of disobedi- 
ence,” “sons of men,” “ children of wrath,” “children of light.” 
I can find no Pauline parallels except 1 Thess 5° “sons 


1See Vocabulary, s.v. 

* At this point Dr. Moulton’s MS ends abruptly in the middle of a 
sentence. 

2 The rest of this section (8 14) is taken from a paper read by Dr. Moulton 
before the Society of Historical Theology, Oxford, on January 24, 1913; it is 
partly a reply to a valuable critique of his Prolegomena by Mr. G. C. Richards 
in JTS x. 283 ff. 

8 See Proleg.* 246 and Hinl. 127. A unique parallel for this ‘‘ Hebraism”’ is 
provided by P. Ry] ii. 113" (a.p. 133), where Hieracion of Letopolis, beekeeper, 
complains of unjust treatment from persons μὴ ἔχοντας πᾶν πρᾶγμα πρὸς ἐμέ : the 
document is very ungrammatical, but shows no marks of Semitic nationality in 
the writer. 


§ 14] INTRODUCTION. 23 


of light and sons of day,” and 2 Thess 25 “son of perdition.” 
The first of these Deissmann regards as a quoted Logion, and 
the second as a quotation from the LXX. In Col 36 “sons 
of disobedience” is interpolated ; and the phrase “ the children 
of the promise” in Gal 4%, Rom 98 is taken out of this 
category altogether by the context. Here, then, is another 
secondary Semitism from which Paul was quite free, except 
when he wrote Hphesians, or (if so preferred) unless he wrote 
Ephesians! The same absence of Semitism comes out for 
Paul by other tests. Ov μή in the NT is characteristic of 
Logia and OT quotations, both, of course, admitting the 
suspicion of “translation Greek,” with the consequence that 
the locution need no more be emphatic than when in the 
LXX it will alternate with ov in one verse as a rendering of 
NS. In Paul it occurs only four times (with two LXX 
quotations to be added), and in all four the emphasis is 
unmistakable, making his use identical with that of classical 
and Hellenistic Greek. Then ἰδού used freely is a natural 
product of Semitic thought. I have compared Fluellen’s 
“look you” as a mark of a Welshman talking English. 
Even the excellent Greek of the Epistle of James may show 
relics of the writer’s Semitic mother-tongue in the frequency 
of ἐδού, as in the isolated προσευχῇ προσηύξατο. Paul uses 
ἰδού only eight times (and once in a quotation), and never 
has a trace of James’s other Semitism, unless Eph 5° is 
rightly thus read? and comes from Paul’s hand. Both ἐδού 
and ἐδέ (semel) are used by him with the classical nwance and 
with normal Greek frequency. The participle with ἦν, etc., 
is probably an Aramaism sometimes in translated books, 
however justifiable as Greek: when Paul uses it, we can 
trace the same force which it has in classical writings. One 
more example may be named—the curious ἐν ῥάβδῳ ἔλθω 
πρὸς ὑμᾶς (1 Cor 474), which even Deissmann had to explain 
away, until Ptolemaic papyri, linked with Lucian, showed 
that ἐν, meaning “armed with,’ was good vernacular 
Greek. 


1 For an interesting conjecture as to the authorship of Ephesians and its 
relation to Colossians, see Dr. Moulton’s popular lectures, rom Egyptian 
Rubbish-heaps, 59 ff. (London: C. H. Kelly. 1916). 

2 See Proleg.® 245, Hint. 119. 3 See Vocab. s.v. ἐν. 


24 A GRAMMAR OF NEW TESTAMENT GREEK. [8 15 


§ 15. When we pass from Paul to that noble work which 
came from the pen of some unknown member of his circle 
we are met by the striking paradox that a letter “to 
Hebrews ” is written by some one who knew no Hebrew, and 
used the Greek Bible alone. It is hardly necessary for our 
present purpose to discuss the structure of this Epistle. 
Whether it was originally an epistle or a treatise} it has a 
literary flavour that distinguishes it from any other book in 
the NT Canon. “Alike in form and contents this epistle 
strives to rise from the stratum in which Christianity had its 
origin towards the higher level of learning and culture.” 3 
We have already 8 referred to the author’s sensitive ear for 
the rhythm of words, and have looked in that direction rather 
than to any elaborate system of rhetoric for the true ex- 
planation of what must strike even the casual reader as a 
distinctive feature of this book. Blass was on surer ground 
when he pointed out another characteristic,—viz. a general 
avoidance of the harsher kinds of hiatus between successive 
words. This would probably be almost instinctive in any one 
who had received a good Greek education, to whom ἐλέγετο 
αὐτῷ would have sounded harsh, much as a word like “ idea” 
sounds harsh in English when followed by a vowel in rapid 
speech. Familiarity with some of the niceties of classical 
syntax may be traced in the exact significance of the tenses, 
in the freer and more skilful use of particles and conjunctions, 
and in the more complex structure of the sentence as compared 
with the other NT writings. Yet his skilful mastery of 
language never betrays the writer into artificiality, nor is bis 
sonorous vocabulary allowed to weary us with the excessive 
use of heavy compounds. Dr. Nairne® calls our attention 
to the “sudden touch of conversational audacity” which 
introduces such a word as πηλίκος into the majestic descrip- 
tion of Melchizedek (74). So flexible a Greek style might 
seem to entitle its possessor to unconditional exemption from 
any examination into his Semitic connexions. There is no 
reason to suspect him of acquaintance with either Aramaic 


1 See Deissmann, BS 49 f., and Moffatt, Zntrod. 428 ff. 

2 Deissmann, LAE 237. 3 Supra, p. 6 n.8 4 CBE 482. 

5 The Ep. to the Hebrews, in CGT, cli. His chapter on The Style of the 
Epistle abounds in illustrations of the characteristics of the author’s Greek. 


§ 15-16] INTRODUCTION. 20 


or Hebrew. Nevertheless one fact necessitates the inquiry. 
This Epistle is steeped in the language of the LXX, and quotes 
from it even more copiously than does St. Paul. It is there- 
fore not surprising to find a formidable list of Semitisms in 
the arraignment. Three of them '—the predicative use of εἰς, 
ov μή, and a violent use of the participle standing by itself 
in the genitive absolute—need not detain us, as they occur 
in a direct citation from the LXX in chap. 8. A similar 
defence might secure speedy acquittal on another count, for 
the phrase ἐν αἵματι διαθήκης (137) is certainly based on 
Zech 91, It should be counted to him for righteousness, 
however, that in his use of such a Hebraism the author 
transcends Semitic idiom.” The presence of a “secondary 
Semitism ” must be acknowledged in ἐν τῷ λέγεσθαι (315). 
Dalman’s claim was denied on the strength of the frequent 
occurrence of the locution in Thucydides, Plato and 
Xenophon; but in view of Dr. E. A. Abbott’s acute criticism, 
that the instances cited must convey the sense of “ during,” 
we have transferred this “Hebraism” to the category of 
“possible but unidiomatic Greek.”* The same chapter 
furnishes another instance of the pervasive influence of the 
phraseology of the LXX. Καρδία πονηρὰ ἀπιστίας (31%) 
can be easily paralleled from Sophocles, as was shown in 
Proleg. 74. But its subject-matter leaves us in little doubt 
that Biblical associations prompted this rather overstrained 
use in prose of the poetical genitive of definition. 


§ 16. In the linguistic conditions of Palestine we have 
already found a clue to the remarkably free Greek of a 
group of writings traditionally ascribed to three members of 
our Lord’s own circle, 1 Peter, James and Jude. When, 
therefore, the question is asked * about the first, “ Is it credible 
that a Galilean fisherman who left out his H’s (Mt 267) 
should after middle life, and in the midst of absorbing occupa- 


1 See, further, Proleg. 72, 74, 187. 

2 Cf. Nairne, ib. exlvii, ‘‘ He will adopt a rude Hebraic use of the preposi- 
tion ἐν, and by careful context fill it with significance, as in 1! ἐν τ. προφήταις 
... & vig, 10 ἐν ᾧ θελήματι, 13°F, where notice how ἐν αἵματι and ἐν ἡμῖν 
explain one another.” 

3 Proleg.® 249, Hinl. 341. 

4 Simcox, The Writers of the New Testament, 68, 


26 A GRAMMAR OF NEW TESTAMENT GREEK. [816 


tion, have learnt to write scholarly Greek like this?” the 
answer is by no means a foregone conclusion. Without 
denying the possibility that this “open letter” owes its 
mastery of idiom to the practised pen of Silvanus, we can yet 
argue that Peter’s Greek may well have been better than his 
Aramaic. A provincial brogue of Aramaic that attracted 
attention in the metropolis of Judaism does not necessarily 
imply defective culture. Moreover, in the thirty-five years 
that lie between the Crucifixion and the probable date of 
this letter, Greek rather than Aramaic would be the tongue 
in which Peter conversed with the Hellenist Jews of Jerusalem 
and Antioch, and the LXX would of necessity be the Bible 
used in all his missionary work abroad. Now it is evident 
that the author of 7 Peter was steeped in the language of the 
LXX. This appears in the number of direct quotations, 
still more in the reminiscences of LXX phraseology which are 
woven into the fabric of his style. In view of the rhythm 
and balance of sentence, the copiousness of vocabulary, and the 
management of tenses and prepositions to which Bp. Chase * 
has drawn attention, it is interesting to notice points of 
contact which the same scholar has indicated between this 
Epistle and the non-Hebraic and literary books of the 
Apocrypha, such as Wisdom, 2 Mac and 4 Mac. From 
Semitisms this Epistle is singularly free, if we exclude from 
consideration the plentiful sprinkling of phrases and citations 
from the LXX. A secondary Semitism may be allowed in 113 
(τέκνα ὑπακοῆς)" and no doubt the OT is responsible for 
such a word as ἀπροσωπολήμπτως (11). The influence of 
the LXX may be traced in such words as ἀναστροφή, 
ἀναστρέφομαι and παρεπίδημος, but the evidence from 
papyri and inscriptions removes them from the old class of 
“ Hebraisms of Vocabulary.” 4 

The bilingual birthright of the Galilean may also account 
for the paradox that “the letter of that specially Jewish 


1 See Dalman, Words, 80. 

2 Hastings, DB iii. 781 f. Mayor (Comm. on Jude and 2 Pet.) even says, 
‘*Perhaps no other book of the NT has such a sustained stateliness of rhythm 
as 1 Feb. 

3 Supra, pp. 22f. Cf. also Deissmann, BS 163 ἢ. 

4 Deissmann, BS 88, 149, 194. Vocabulary, s.vv. 


§ 16] INTRODUCTION. a7 


apostle, St. James, is perhaps the best Greek in the New 
Testament.”! We need not linger over the theory of an 
Aramaic original.2 The writer’s fondness for paronomasia 
and alliteration do not suggest the hand of a translator; and 
whilst the erisp vivacity of James offers a striking contrast 
to the rather long and well-balanced sentences of / Peter, 
such constructions as ἄγε νῦν, ἔοικεν, χρή, ἀπαρχή Tis, are 
evidence that his style takes high rank in the Greek of the 
NT. Two traces of the author’s Semitic mother-tongue have 
already been mentioned (p. 23). We must also notice an 
overstrain in the use of the genitive of definition in such 
phrases as TO πρόσωπον τῆς γενέσεως αὐτοῦ and ἀκροατὴς 
ἐπιλησμονῆς. Instrumental ἐν in 3° was formerly counted a 
Hebraism, so were the aorists in 14, But the publication of 
the Tebtunis Papyri closed the controversy about the former 
(ef. supra, p. 23), as the weighty judgement of Hort * may 
be held to have settled the other question in favour of the 
gnomic aorist. 

The little letter that bears the name of Jude is chiefly 
remarkable for the wealth of its vocabulary, derived in part 
from the LXX, the pseudepigrapha and the Pauline Epistles. 
Its fondness for sonorous words might seem to suggest a 
wider acquaintance with literature. But writers of the Κοινή, 
“embodying older strata of the language, would suffice to 
supply him with his vocabulary.”* Less flexible than / Peter 
in syntactical structure, Jude also lacks the epigrammatic 
succinctness of James. Vigorous and descriptive he certainly 
is; and the tendency to triple expression is a well-marked 
feature of his style. Mayor’s exhaustive analysis of his 
grammar shows that the author was quite at home in 
Hellenistic idiom. Indeed the critical microscope fails to 
detect a genuine Semitism in the 24 verses. 

Another writer calls for brief mention with this 
Palestinian group on the ground of literary indebtedness 
rather than from considerations of style. The “ Atticism” 
of 2 Peter has been mentioned in an earlier section (§ 5), 


1 Salmon, Jntrod.* 139. 

2 Bp. John Wordsworth’s a priori arguments have been fully answered by 
Mayor (Comm.? ccexxxii ff. ). 

> Comm. on 1 Pet. 96, * Chase, Hastings’ DB ii. 801. 


28 A GRAMMAR OF NEW TESTAMENT GREEK. [§ 16-17 


which explained the unique character of its Greek as an 
artificial dialect of high-sounding words learnt from rhetoricians 
or books and employed with the uneasy touch of one who 
acquired the language in later life. It is significant that 
this Epistle has not a single quotation from the OT and but 
five uncertain allusions. This absence of the phraseology of 
the LXX may account for its freedom from the slightly 
Semitic colouring that we have noticed in the admirable 
Greek of the other Epistles in this group. In only two places 
is there any real approach to Semitism. “Ev ἐμπαιυγμονῇ 
ἐμπαῖκται may belong to the same class as James's 
προσευχῇ προσηύξατο. But in view of the slight impres- 
sion which the LXX has left on his diction, and bearing in 
mind Dalman’s caution,’ it seems more reasonable to explain 
this locution as an instance of the author’s tendency to 
reduplication. Κατάρας τέκνα claims kinship with that 
familiar genitive of definition, though a quasi-classical turn 
is given to the phrase by the inverted order of the words, 


§ 17. Starting from Luke we have fetched a wide com- 
pass and must now return to the other three Gospels. 
There can, of course, be no question that translation-Greek 
occurs in Mark and those parts of the Synoptic Gospels 
which reproduce “Q.” Enough has already been said to 
indicate that real Aramaism may be allowed ungrudgingly in 
those parts of the NT which are virtually translated from 
Aramaic oral or written sources. Wellhausen’s brilliant 
investigation serves to remind us of the need of keeping in view 
the distinction made above between secondary Semitisms and 
Semitisms pure and simple. For like other Semitic specialists, 
that lamented scholar was perhaps sometimes in danger of 
recognising foreign idiom where a Greek reader of the book 
would never suspect anything wrong. Thus in Lk 14:8 ἀπὸ 
μιᾶς is said to be the Aramaic min ch’da, which he calls “a 
thumping Aramaism.” But why call it an Aramaism when 
the phrase was a very idiomatic expression in Luke’s own 
language? It is merely a case of coincidence between the 


1 Words, 34. ‘*The Hebrew mode of emphasising the finite verb by adding 
its infinitive or cognate substantive . . . is in the Palestine Aramaic of the 
Jews—apart from the Targums—quite unknown,” 


ὃ 17] INTRODUCTION. 59 


idioms of two languages; and while Mark perhaps might have 
been ignorant of it as Greek, and was capable of employing it 
as a literal translation of the Aramaic, Luke simply could not 
have used it as other than a normal Greek term. Although 
the same instinct for Greek style cannot be claimed for 
Matthew, the First Gospel betrays Semitic authorship only 
in its range of ideas and its sympathetic understanding of the 
Jewish-Christian point of view. The language, on the other 
hand, is a correct if rather colourless Greek which avoids the 
vulgar forms without displaying a mastery of the literary 
syntax. The Hebraisms which in Luke express a literary 
feeling that formed itself on ancient models are noticeably 
wanting from Matthew’s narrative. Instances have already 
been given of his editorial revision where Luke preferred to 
sacrifice style in favour of fidelity to the original source. The 
same freedom is shown in his treatment of Mark,! so much so 
indeed that it has been said? that “Matthew grecises Mark.” 
One or two exceptions, however, may be noticed because of 
their bearing on the question of Semitisms. The phrase 
πέμψας διά is unquestionably the right reading in 11%, and 
Wellhausen points out that a ποιοὶ in Aramaic as in Hebrew 
is a regular phrase for sending a message. In the Lucan 
parallel διά has become δύο τινάς, from which it seems to 
follow that Luke misread 8 the literal διά of his source, which 
is preserved in Matthew. Incidentally we have here strong 
evidence in favour of the Greek basis in the common source. 
Similar testimony is borne by the notable logion in Mt 107&* 
where, in addition to the sequence of words in v.” and 
the whole of v., the Lucan parallel gives us also 
Matthew’s ἐπὶ τῶν δωμάτων, a good Hellenistic phrase ; 
φοβηθῆναι ἀπό, a very marked piece of translation-Greek ; 
and ὁμολογεῖν ἐν ἐμοί, Deissmann*‘ describes this last as a 
translation made “ with a painful scrupulousness coming near 
to a pedantry of interpretation.” Even the author of the 
Apocalypse gives this logion in a better Greek form (ef. 


1 For instances see W. C. Allen, ExpT xiii. 328 f., and Comm. on Matt. 
ΕΣ fr 

2 This is Wellhausen’s summary (Hin/.! 35) of Zahn’s argument as quoted 
by Wernle, Die synoptische Frage (1899, vii, viii and 120). But see Zahn, 
Introd. (E.T.) ii. 576, 591-2. 

8. See § 18. * In Christo, 60, quoted in Proleg. 104. 


30 A GRAMMAR OF NEW TESTAMENT GREEK. ([§ 17 


Rev 3°). Turning to Matthew’s treatment of his other chief 
source we note another apparent exception to the general 
rule. In the lawyer’s question (Mt 22%*) ποία ἐντολὴ μεγάλη 
ἐν τῷ νόμῳ; Mark has not μεγάλη but πρώτη πάντων, a 
thoroughly vernacular phrase. Matthew’s form is quite 
clearly a much less idiomatic translation of the Aramaic 
(which has no degrees of comparison), and it falls in with 
various indications that he is here drawing upon other material, 
quite possibly derived from Q. This literal translation of 
the Aramaic would,be very natural when we bear in mind 
the obsolescence of μέγιστος. Another instance of translation- 
Greek which Wellhausen notices is the impersonal use of the 
3 plur. active in place of the passive. This is common to all 
the Syunoptists, but may be mentioned here because of two 
occurrences in Matthew which raise points of interest in 
Synoptic criticism. In 1% καλέσουσιν replaces the more 
natural καλέσεις of the LXX and suggests an Aramaic trans- 
lation from the Hebrew current in Matthew’s time as part of 
a collection of testimonia” Once again in 5” the Matthean 
οὐδὲ καίουσιν καὶ τιθέασιν preserves an Aramaism where Luke 
employs the more idiomatic οὐδεὶς ἅψας τίθησιν. The fact 
that this logion is a doublet in Luke and that in the Marcan 
passage the Aramaism is absent, seems to show that the say- 
ing was current in more than one form, which would account 
for the divergence without crediting it, with Harnack, to Luke’s 
stylistic improvement of Q. In many of these instances given 
by Wellhausen,? although the active use of the verb is quite per- 
missible Greek,‘ it must be allowed that the passive would prob- 
ably have been used but for the influence of a Semitic original. 

The language of our Second Gospel shows a very marked 
deficiency in culture on the part of the Jerusalemite Mark, 
who seems to have a foreign idiom perpetually behind his 
Greek. “ Peter's former interpreter ”—for so we must render 
Papias’s phrase °—had been a ὑπηρέτης or “ minister of the 


1See Proleg. 78, and cf. the frequent introduction of μέγας μέγας in early 
papyri (as P Tebt i. 63° (ii/B.c.)) ; so Hinl. 122, 124. 

2So A. H. McNeile, Comm. in loc. 

ὃ Tbid. 25 f. 4Cf. Proleg.? 58 f., also Hint. 87. 

5 Μᾶρκος μὲν ἑρμηνευτὴς Πέτρου γενόμενος (or ‘‘ having been Peter’s interpre- 
ter”); cf. Vocabulary, s.v. γίνομαι. 


§ 17-18] INTRODUCTION. 31 


word,” that is, a teacher or catechist who accompanied an 
apostle on his missionary tour for the purpose of instructing 
inquirers in the outlines of the life and teaching of that Jesus 
whom they preached. We may repeat here what has been said 
elsewhere : 1 “There can be no question that the catechetical 
lessons on which the written Gospel was ultimately based, 
were given first in Aramaic; and they may well have become 
so fixed in that form that when their author transferred 
them to Greek they retained ubiquitous marks of too literal 
translation. It is of great critical importance to observe how 
these Aramaisms of translation were progressively smoothed 
away. Wellhausen shows that D has most of them and B 
distinctly less. Unless this is due (as Bishop Chase argued) 
to a Syriac infection in D, we have here a most important 
source of evidence as to the origin of the Western Text, of 
which in this respect the ‘Neutral’ becomes a revision. As 
has been noted already, there is plenty of revision of Mark’s 
Aramaism to be seen in Matthew and Luke. In a consider- 
able number of little points these Evangelists coincide in 
their amendments, a fact well explained by Dr. Sanday’s 
suggestion? that the text of Mark had been polished by a 
cultured scribe before it reached them: our Mark descends 
from the unrevised form. Mark’s Semitisms . . . are hardly 
ever really barbarous Greek, though his extremely vernacular 
language often makes us think so, until we read the less 
educated papyri. Generally we recognise them by their over- 
use of a possible though uncommon idiom which happens to 
agree with Aramaic.” 

A singularly neat instance came to light in a Berlin 
papyrus*® to confute those who would describe βλέπειν ἀπό 
(Mk 8” etc.) as a rank Hebraism: ὡς ἂν πάντες καὶ σὺ βλέπε 
σατὸν ἀπὸ τῶν ᾿Ιουδαίων. Surely it was no Jew who gave 
this warning to his friend ! } 


§ 18. The Fourth Gospel and the Johannine Epistles 
(which, on every consideration of style, form with it a 
literary unity) are the work of a writer to whom Greek was 


1 Camb. Bibl. Essays, 491. 2 Oxford Studies, 21. 
ὃ BGU iv. 1079% (a.p. 41) (= Selections, p. 40). 


39 A GRAMMAR OF NEW TESTAMENT GREEK. [ὃ 14 


evidently no mother-tongue. We infer this from the exces- 
sive simplicity of the style and its poverty of idiom, not from 
any grammatical aberrations. The conditions lend support 
to the theory that the author was brought up in Jerusalem 
(see p. 13). But in spite of certain superficial indications 
that point that way the style is not Semitic. Two such 
indications in the Gospel may be briefly noticed,—the pre- 
vailing use of parataxis and the priority of the verb in the 
sentence. As for the former, parataxis predominates to 
such an extent that we instinctively recognise an editorial 
hand in the flowing periods that form the prologue to the 
Passion narrative (13!+*). Yet it is impossible to claim that 
the incessant co-ordination of simple sentences by καί is a 
hall-mark of Semitism after studying Deissmann’s parallel} 
between the narrative in John 97" and a Roman inscription 
of the time of the Antonines giving an account of the 
marvellous cures wrought by Asclepios. A stronger argu- 
ment can be based on the arrangement of words, for, as a 
rule, in the Johannine writings the verb stands first and the 
subject follows, and it is tempting to trace the Semitic genius 
of language in such an order. Even upon this subject it is 
well to suspend judgement until careful statistical investigation 
on the lines of Kieckers’s important monograph ? has provided 
us with material for a wider induction. Meanwhile we do 
well to observe other significant features in the order of 
words. Wellhausen * attributes the precedence given to the 
verb to imitation of the Biblical style as best fitted to the 
subject-matter. He even discovers in the solemn rhythm a 
self-conscious, sacerdotal language, to be compared with the 
pedantry of the Priestly Code in the Pentateuch. Whatever 
we may think of this judgement we must accept his verdict 
that the position of words in general is unSemitic. This 
can be seen best in the tendency to remove the subject to the 
very end of the sentence (cf. 29, 63, 1 888. 19%), and in the 
position of the dependent genitive in phrases like αὐτοῦ οἱ 
μαθηταί, δύο ἀνθρώπων ἡ μαρτυρία, etc. The great Semitist 
discovers no trace of the construct state in John, and stays 
1 LAE 131. 


2 Die Stellung des Verbs im Griechischen (Strassburg, 1911). 
3 Das Evangelium Johannis, 1383-146. 


§ 18-19] INTRODUCTION. 99 


his hand when he finds an occasional casus pendens followed 
by a resumptive pronoun. The linguistic evidence all goes 
to show that the author of the Fourth Gospel was a man who, 
while cultured to the last degree, wrote Greek after the 
fashion of men of quite elementary attainment. His uneasy 
movement in the region of unfamiliar idiom is never suffered 
to betray him into a breach of the laws of grammar. 


§ 19. Very different are the phenomena that meet us in 
that marvellous book which so fittingly closes the canon of 
the NT. The Greek of the Apocalypse differs in an 
extraordinary degree from that of the Fourth Gospel. Not 
only does it display a greater freedom in copiousness of 
vocabulary and elaborate phraseology ; it is simply defiant of 
the restraints of grammar. Various attempts have been 
made to secure the traditional unity of authorship with the 
Gospel. But even Hort’s strong argument for an early date, 
leaving thirty years in which the author could improve his 
Greek before writing the Gospel, falls short of a solution. 
Modern criticism has little to add to the penetrating analysis 
which Dionysius of Alexandria furnished in the middle of the 
third century. Speaking with the authority of one to whom 
Greek was a native tongue, this critic discerned a difference 
not of degree but of kind. Even if the decision against an 
early date were not fairly unanimous, it would still tax our 
ingenuity to bridge the chasm between the unchartered 
liberty of Revelation and the austere simplicity of the Gospel 
according to St. John. The Apocalypse, in the contrast it 
affords between wealth of diction and grammatical solecisms, 
suggests an author who had used Greek all his life as a 
second language and never from choice. His seeming 
indifference to the rules of concord can be readily understood 
by Englishmen who stumble over the genders of French and 
German after speaking a language unburdened with this 
useless survival. A fresh impetus has been given to the 
study of this strange dialect by Dr. R. H. Charles, who has 
shown in his Studies in the Apocalypse! that many of its 


1 And with greater fulness of detail in his two-volume commentary in the 
Int. Crit. Com. from which Canon Charles has kindly shown me extracts in 
the proof stage. 

VOL. II.—3 


94 A GRAMMAR OF NEW TESTAMENT GREEK. [819 


mannerisms are due to the literal transference of Semitic 
idioms. A striking illustration (found seven times in Rev) 
is the co-ordination of the participle in one of the oblique 
cases and the finite verb, eg. 2% τοὺς λέγοντας ἑαυτοὺς 
ἀποστόλους καὶ οὐκ εἰσίν. But while the book abounds in 
translation-Greek and bears constantly the imprint of the 
author’s Semitic mind, it is easy to go too far in attributing 
all its peculiarities of grammar and idiom to the influence of 
the LXX when a Hebrew or Aramaic source is not in 
question. Thus Moffatt’s treatment of the subject! finds a 
useful counterpoise in Radermacher’s? judgement. After all 
the author was capable of writing a vigorous though irregular 
Greek with a very free pen and, as Dean Armitage Robinson 
has pointed out,? “the Greek in which he expressed himself 
was more like the Greek of the Egyptian papyri and of 
inscriptions found in various parts of the Graeco-Roman 
world.” The very blunders in concord do not imply 
ignorance in the ordinary sense; “it is familiarity with a 
relaxed standard of speech, such as we find often enough in 
the professional letter-writers who indited the petitions 
and private correspondence of the peasants of the Fayim.” 
Perhaps it was but fitting that the weird melodies and 
daring harmonies in which the seer of Patmos gave utterance 
to the things which he had seen “which must shortly come 
to pass,” should speak to us now in the haunting cadences 
of Jewish apocalyptic, and again in the popular idiom of the 
Graeco-Roman world. 


1 Introd. 501. 

2 Neutestamentliche Grammatik, 87: ‘‘ Was die Apokalypse, und zwar sie 
allein unter den Schriften des Neuen Testaments, an entsprechenden Fallen 
zeigt, hat also nicht mehr als Solocismus zu gelten und darf schwerlich als 


sklavische Nachbildung eines hebriaischen Originaltextes erklirt werden.” 
5.01. x. 9. 


PART I. 


SOUNDS AND WRITING. 


Scene ἘΦ ΒΞ Μ'ὶ SSO nA ΞΡ nA τ ΘΝ ΞΡ ΞΡ ἘΞ 


PAR EG! 


PHONOLOGY AND WRITING. 


§ 20. The Greek Alphabet in the Hellenistic period had 
twenty-four letters :— 


Form. ens Name. 
2 3 In Greek. In English. 
δ α a ἄλφα Alpha 
B β b βῆτα Beta 
r γ g γάμμα Gamma 
A ὃ 4 δέλτα Delta 
€ €, € ὅ εἶ, later ἔ Epsilon 
Ζ C Ζ ζῆτα Zeta 
H ἢ 8 ἦτα Eta 
8 0,3 th θῆτα Theta 
Ι ι 1 ἰῶτα Tota 
K K k (ec) κάππα Kappa 
λ λ l λά(μ)βδα Lambda 
M μ ΠῚ μῦ Mu 
N ν n νῦ Nu 
Ζ ξ x ξεῖ ΧΙ 
ο 0 ῦ ov, later o Omicron 
T π ip πεῖ Pi 
Ρ ρ ie ῥῶ Rho 
G Gas 5 σίγμα Sigma 
τ τ t ταῦ Tau 
Y υ u (y) ὖ | Upsilon 
b | φ ph φεῖ | Phi 
X x kh (ch) χεῖ Chi 
y ψ ps Wee Psi 
w w 0 ὦ Omega 


37 


38 A GRAMMAR OF NEW TESTAMENT GREEK. [§ 20-21 


Notes. 


1. The first column represents the printed form of the capital letters, 
based on the alphabet (ra Ἰωνικὰ γράμματα) which from iv/B.c. was 
generally used in inscriptions. For the history of the forms see §§ 21 f. 

2. In the second column stands the alphabet of the oldest uncial MSS, 
as printed by WH in citations from the OT. The third column is the 
ordinary alphabet of modern printed books, based on that which the 
early printers derived from “cursive” or “minuscule” MSS. The 
alternative forms here given are indifferent except in the case of σ΄ s, the 
latter of which is used at the end of words only. Some print it also at 
the end of a preposition or adverb compounded with another word 
(προςφέρω etc.); but the historical justification of this form does not 
apply to any position other than the actual end of a word.! 

3. The transliteration column shows the value of the letters as 
evidenced by the form they take in contemporary Latin : an exception is 
made with x, v, x, which in Latin take the bracketed form only. The 
pronunciation of the letters will be treated summarily below S§ 23 ff., 
and in detail under the Orthography. 

4. The names of the letters are given in English according to their 
conventional forms, many of which, however, are unwarranted by early 
usage. The addition of ψιλόν to the names ἔ and ὖ is a late misunder- 
standing: Byzantine grammarians, giving rules for writing a or e, οἱ or v 
(which were not distinguished in sound), would say that a particular 
word was written διὰ τὸ ε ψιλοῦ, “ with a simple ec,” as distinguished from 
a diphthong a. “Little o” and “big o” are names dating from the 
Byzantine period, when the two letters were pronounced alike. The 
name for ἃ is better attested as Labda than as Lambda. 


§ 21. For the history of the alphabet, a 
long and complex subject, reference may be 
made to special works: see literature in art. 
“Writing,” Hne. Brit. (Ὁ. Giles). It must only be mentioned 
here that the alphabet (in Greek γράμματα, of the forms 
of the letters, στοιχεῖα, of the sounds) came ἴῃ pre- 
historic times? into Greece from Phoenicia. The date 


History of 
the Alphabet. 


1 In the earlier printed books we find the compendia $=o7, 8=ov. 

2 The event is already covered with a myth, Cadmus, the “ eastern” (07p), 
being credited with the καδμήϊα γράμματα (Herod. v. 58), which are, however, 
Φοινικήϊα in the same context and elsewhere. Noideke (Beitr. z. semit. 
Sprachwiss., 1904, 124-136) rejects the suggestion that Aramaic influence is 
traceable in the names ἄλφα and others: the forms are only due to Greek 
modification. (I owe the reference to Prof. Hogg.) Nestle (Philologus, 1900, 
476 f.) says the theory is as old as the sixteenth century. Dr. A. J. Evans 
has shown that the Phoenicians themselves derived the alphabet from the 
prehistoric Cretan script of the newly discovered Minoan inscriptions ; see his 
Scripta Minoa (1909), 86 ff. 


§ 21-22] SOUNDS AND WRITING. 39 


may possibly be prior to the composition of the oldest parts 
of the Jliad, but certainty on this famous question can hardly 
be expected. It is a very striking fact that contact with 
the Semites should have occurred before the dawn, and after 
the sunset of classical Greek literature, and hardly at all in 
the interval. The letters seem to have been adapted to 
some extent independently by different Greek communities. 
But we find in all alike the central principle which betrays 
Greek genius at work even when, for once, it was borrowing 
and not inventing. An alphabet without vowels would 
have been peculiarly useless for expressing Greek words. 
Accordingly superfluous consonants were adapted to new 
uses: δὲ became adda, nm supplied εὖ, n Ara (in earlier times 
h, as in the local alphabet with which Latin was ultimately 
written), was ἰῶτα, y was οὗ. Later invention produced 
ὃ and ὦ. In several Greek dialects ἡ survived as F (Bad, or 
Siyaupa, from its form); but this sound disappeared in Attic 
before the birth of its literature, and there is no trace of it 
left in the Κοινή. Its sign was still used in numeration : 
see ὃ 70. So also with p («om7a) and y (cdv),! the former 
of which (preserved in the Latin alphabet) still survives in 
the earlier period inscriptions to represent the k-sound 
before o and uv. The other equations may be summarily 
stated: 3=8,3=y, 1=6, 1=6, 0D=6,5=«, S=X, 2) yh, 
1=v,}=7, =p, ¥(?)=0,N=7. In Ὁ adapted as Ewe have 
an application of a useless letter which was at first confined 
to the Asiatic Ionians; but it was their alphabet which 
ultimately established itself in universal use. The last five 
letters of the Greek alphabet are later additions, and of 
these v alone is found in all parts of Greece alike. The 
different value attaching to X in Latin, which owed its 
letters to Chalcis in Euboea, through its colony of Cumae, 
illustrates the independence with which these non-Phoenician 
signs were used in different localities. 


§ 22. We must not tarry here to show how the forms and names of 
the Semitic letters are related to the Greek derivatives, nor how in 
Greece itself these letters and their names varied from place to place and 


1 There is considerable doubt as to the history of the sibilants: see Roberts, 
Greek Epigraphy, i. 9f., where x is identified with σάν. 


40 A GRAMMAR OF NEW TESTAMENT GREEK. [Ἃ8 22-23 


generation to generation, till the final victory of the Ionic alphabet in 
which alone the Common Greek was ever written. Two points only may 
be selected from the history of Greek writing, the development of the 
Breathings, and the change of direction from the retrograde Semitic to 
that which we inherit now. In the pre-Ionic alphabets H was used for 
h; but when it was requisitioned to express ὃ, the first half of it, +, was 
used as an aspirate sign, perhaps as early as iv/B.c. (Blass). The 
Alexandrian grammarians introduced the corresponding 4 to represent 
the voiced sound answering to the breathed h, the glottal catch which 
Semitic languages wrote with &: it is a sound with which every vowel- 
initial opens, if not aspirated. From + and 4 respectively are derived 
the ‘ and ’ which we call rough and smooth breathing (πνεῦμα δασύ 
and πνεῦμα ψιλόν). The other matter, the direction of writing, 
need only be mentioned here because the relation of the Greek to 
the Hebrew writing specially interests students of the Greek Bible. It 
it enough to say that in the earliest Greek inscriptions the writing is 
from right to left ; that this develops into what was called βουστροφηδόν, 
because it “turns” at the end of each line in the opposite direc- 
tion as the “ox” does in ploughing; and that out of this in v/B.c. 
developed the left-to-right style which Greece passed on to Rome, 
and Rome to modern Europe. The three stages may be illustrated 
by short examples. (1) F2O4Q ἐποίε(ι) (Thera, vii/s.c.). (2) EM 
POLEMO! 
MOMAMI@O ἐν πολέμῳ φθίμενον (Attica, vii/B.c.). (3) PAS 
A4OM παῖ Διός (Thera, vii/B.c.). 

On the whole subject see Roberts, Introduction to Greek Epigraphy, 
vol, i. (Cambridge, 1887). 


Classification of Sounds. 


§ 23. Greek sounds are thus classified 
for the Hellenistic period. 

(a) Vowels :—a, t, v, long or short; ¢, 0, (short only); 
η, ὦ, (long only). 

(Ὁ) Diphthongs :—au, εἰ, ot, vt (short); @, 7, @ (long); av, 


ev, ov (short); nu(long). For the vowels and diphthongs see 

(c) Mutes.—These sounds are divided in two ways: (1) 
according to the point of articulation, (2) according to the 
presence or absence of voice, 1.6. the tension of the vocal 
chords, and of aspiration, te. the accompaniment of an 
h-sound. By the first, sounds are classed as dabials, made 
with the lips, dentals, made with the tip of the tongue 
against the teeth, and gutturals or palatals, made with the back 
of the tongue against the palate. By the second, they are 


Sounds. 


§ 23-24] SOUNDS AND WRITING. 41 


breathed or hard, voiced or soft, and aspirate, the last class being 
breathed (hard) as well. The classification is as follows :— 


Breathed. Voiced. Aspirate. 
Labial π B φ 
Dental τ ὃ θ 
Guttural K y x 


There are other names used for breathed and voiced, such as surd and 
sonant, tenues and mediae, smooth and middle: Goodwin, who employs the 
last-named pair, calls the aspirates rough. 


(d) Continuous, or Semivocalic consonants. These sounds 
differ from the Mutes, or stoppéd sounds, in that they are 
capable of prolongation, and may even become vowels. They 
are classed thus. Spirants result from relaxing slightly the 
contact which produces a mute. Thus the position of ¢, 
if the tongue is held loosely to allow breath to pass, produces 
our English th. Nearly the same position, with a groove 
along the tongue, results in o, the only spirant represented 
in the Greek alphabet (see however § 43): the breathings 
“and ‘ belong to the same class. The labial spirant F, our 
w, is obsolete in Hellenistic Greek. Nasals are characterised 
by the opening of the nasal passage. They are three in 
Greek, w (labial), ν (dental), and that which is written y 
before gutturals («, y, x, &), the guttural nasal ng (as in 
sing). Inquids are p and δ : the same rather elastic term is 
often used to include w and ν. Both p and ἃ are made with 
the tip of the tongue against the front palate: in X it is at 
rest there, while voice passes on both sides; in p it vibrates, 
while the two sides are stopped. Initial p in earlier Greek 
was breathed, as also is the second element in the doubléd pp: 
it may be questioned whether this was really true for i/A.D. 

This classification does not include the digraphs y, ¢, &, 
which are mere combinations of labial, dental and guttural 
mutes with the sibilant, like our superfluous letter x The 
pronunciation of ¢ however fluctuated considerably: see ὃ 43. 


Pronunciation. 


§ 24. The pronunciation of Greek in the 
Hellenistic period raises a great many diffi- 
cult questions which cannot be discussed 
here. It is probable that considerable differences existed 


Summary of 
Pronunciation. 


42 A GRAMMAR OF NEW TESTAMENT GREEK. [ὃ 24 


between the Greek of Rome and Asia, Hellas and Egypt. 
The pronunciation of i/—iv/a.D. is a matter of great import- 
ance from its bearing on textual criticism. If we could 
delimit the localities affected by certain variations, we should 
have important evidence for the localising of textual types. 
Unhappily our information is too scanty to make this a 
really useful resource. Pronunciation had greatly changed 
since the classical period. As shown already (Prol. 34), 
Hellenistic Greek, though written with Attic orthography, 
sounded much more like Boeotian than Attic. Many of the 
processes had already started which reach their full effects in 
MGr. It does not follow that to pronounce Hellenistic as 
if it were MGr would compensate in accuracy for the in- 
convenience it would cause. For pronouncing Attic of the 
classical period, the MGr system is almost as wide of the 
mark as our English system of reading Greek as if it were 
English—a system which pretends to no advantage but 
convenience. For Hellenistic, it is much closer, but still far 
from exact, as we shall see; and the practical awkwardness 
of blotting out the difference between εἰ, ὁ, ἢ, 7, οι and vu 
would be too great a price to pay for the approximation 
gained. 


Under the head of Vowels, we have first the dis- 
Vowels and 4. tion b 1 PR een re ete 
Diphthongs. tinction between long and short. In classical Greel 
this is felt throughout the language. It is the basis 
of all metrical composition, when combined with the rules by which 
a short vowel before groups of consonants counted as long. Its 
influence upon accentuation will be seen in § 29. The essence of the 
distinction is that a long vowel took the time of two shorts : in technical 
language a short vowel had one mora, a long vowel two. The progres- 
sive disappearance of this distinction is one of the most important 
changes in the language. It established itself very gradually, and locali- 
ties differed widely in their treatment of the several sounds concerned. 
Perhaps Asia Minor was its earliest home: Greece proper was the latest 
to accept it. The main cause of this levelling was the change in the 
character of the accent (see § 29): when stress replaced pitch accent, the 
accented syllable tended to be long, and the unaccented inevitably was 
shortened. We may take the process to have been complete before the 
date of our oldest uncials, though many quantity-distinctions were still 
operative in i/A.D. 
The other general tendency needing mention at this stage is that by 
which the diphthongs as such were destroyed, being replaced by simple 
sounds, This was a tendency which Hellenistic owed to Boeotian 


- 


§ 94] SOUNDS AND WRITING. 43 


phonology : see Prol. 33. The only exceptions were founa in av and ev, 
which tended to consonantise the second element, so that the MGr pro- 
nunciation is av, ev (or af, ef, before breathed consonants). Even 
these last changes were incipient already in our period, and were 
complete before the mass of our MSS were written. 

On the vowel a, long and short, it need only be said that its pro- 
nunciation was the same as in nearly every I.E. language except our own. 
The a was sounded as in father, and the a, so far as it was still dis- 
tinguished, was the same sound pronounced in half the time. Our 


characteristic a (in man, only short) was unknown in Greek. From a, 
in which both lips and throat are wide open, the vowels bifurcate 
along two lines, according as the throat is narrowed or the lips rounded 
and contracted. The former class includes the e-7 vowels, the latter 
the o-u. 

In classical Greek of early times the progressive narrowing of the 
throat aperture produced the series n (long, open), ε (short or long, close), 
c (short or long). When e was long it was written e in the Ionic alphabet : 
it must be carefully distinguished from the genuine diphthongal εἰ, which 
was never written with simple E.! Open e is the sound of French ὃ, our 
ein there; close ὁ is French ¢, which we do not possess—our common a 
in day, datsy, date, is the same sound with an ὁ “glide” making it 
diphthongal. In the Hellenistic period there were many changes going 
on, and in different directions according to locality. H in many places 
is still open e, especially in the area once occupied by Ionic. In inscrip- 
tions of Asia Minor we very often find ἡ confused with ε (short), which 
was by this time probably not as close as it was in Attic: ε isa medium 
e now. But in the Greek of Hellas itself it would seem that η had 
become closer than ε, as in the Boeotian of the earlier age; and the 
difference is reflected to-day. In the Pontic-Cappadocian MGr ἡ is 
frequently ε, as πεγάδι from πηγή ; and such general ΜΟῚ words as 
στέκω, from Hellenistic στήκω, may have originated in districts where 
this pronunciation was the rule. But in continental MGr ἡ is now a 
simple ὦ : this change was complete before vi/a.pD. Finally 1, with which 
εἰ was now wholly convertible even when short, was the closest vowel 
of all: note that in transliterating Latin, where z was open like our 7 
in kin, e was often used instead of ε. 

By rounding the lips and progressively contracting the aperture there 
came in early Greek the corresponding series w (long, open), o (short or 
long, close), and in one or two dialects, as Boeotian, v short or long (as 
in full, fool). In Attic however, and most dialects, v had become ἐΐ 
(German, like the French u), which is pronounced by simultaneously 
rounding and contracting the lips and narrowing the throat aperture. 
This was still the sound generally in our period. Meanwhile the close 
long o, written ov—as in the case of εἰ, to be distinguished from the 


1 Thus εἶμι tbo, where « is radical, was always ΕἾΜ] ; εἰμί swm, where εἰ is 
simply ε lengthened by compensation for the loss of σ, was written EMI in 
Attic before the archonship of Euclides (403 B.c.). 


44 A GRAMMAR OF NEW TESTAMENT GREEK. [ὃ 24-25 


genuine diphthong ov—had become so close as to answer entirely to our 
own oo. In Hellenistic o no longer seems to differ from ὦ ἴῃ the original 
way, as the o in mote differs from that in more. The original open 
vowel, as in the case of ἡ, has become closer: as ἡ approximates to 1, 
so ὦ ultimately does to wu. In i/a.D., in many parts of the Greek-speak- 
ing world, there was little perceptible difference between ο, and ov, a 
fact which leaves its traces on our texts. 


The Diphthongs were largely monophthongs by the time with which 
we are concerned. Ac seems to have become entirely equivalent to ε, 
(but see ὃ 36). Ex and ., even ἵ, are constantly confused ; and in both 
these diphthongs the MS tradition is valuable only in so far as it may 
preserve a historical difference affecting the sense—if an unlettered scribe 
supplied a link in the chain, its value for us is gone, and we can write a 
or ¢, εἰ or «, according to our own preference. Οὐ was in classical Attic 
not very different from our οὐ; but it passed through ὃ (German, as 
French eu) to ἐΐ, like v, with which it ultimately coalesced. Illiterate 
papyri of i/a.D. and even earlier show this confusion ; but outside Egypt 
the sounds were distinct for generations after this time. Yu, in vids, ὀργυιά 
and perfect participles feminine, maintains itself against the literary v: 
it was presumably still ay. The v-diphthongs with long vowels as their 
first element—in which ὁ has since xii/a.D. been subscript—completely 
lost their diphthongal character at an early date. Hux was fused with εἰ 
in Attic; but the indifference with which c is added, especially in i/B.c. 
and i/a.D., to long vowels without justification,’ shows that in the Κοινή 
ἢ was nearer to ἡ than to εἰ, while ᾳ and » were identical with ἃ and o. 
See further § 36. 

In the v-diphthongs (av, ev, ov, nv) the v had from the first the sound 
of simple u. The tendency to consonantise this w in the case of av and 
ev has been already mentioned. Apart from this, the pronunciation of 
av and ev, nv was normally what we should get by running together ah-oo, 
eh-oo respectively. Ov had been for long simply %, sometimes represent- 
ing even w (in full). In the case of av there was a separate tendency to 
slur and finally drop the v when it came before a consonant: a similar 
tendency is observable in the later vernacular Latin. For NT exx. of 
this see Prol. 47. 


§ 25. We pass on to the Consonants, and take first 
the nine Mutes or “stopped” sounds. The breathed π, 
τ, « call for no remark, as they were pronounced very nearly as in English. 
The voiced B, δ, y, had changed considerably from the sounds they had 
in Attic, which were virtually 6,d and g. I was clearly (as partially in 
MGr) the voiced form of ch (Scotch or German), a guttural spirant like 
the German g between vowels (Yage), not far from the English y. 
Similarly 8 and ὃ came to be our v (nearly) and th (as in thou), but the 
change was not completed so early, and it seems to have varied in differ- 
ent districts. The fact that ἡ consonantal is still v in Δαυείδ (all uncials 


Consonants ; 


1 Thus in BGU iii. 883? (ii/A.D.) . . . ὀκτώι, οὐλὴι δακτύλω πρώτωι. 


§ 25] SOUNDS AND WRITING. 45 


which do not use exclusively the abbreviation), and the Latin conson- 
antal u in Silwanus is ov (see ὃ 37), shows that the interlabial v sound of 
8 was not established in the Greek of the NT writers; otherwise the 
sound was so near the w that we should have had a certain amount of 
inconsistency in our documents. (Ata later time Δαβίδ and Δαυείδ would 
be pronounced alike.) The aspirates were during the classical period 
mutes followed by h: our shepherd, hothouse, packhorse give the sounds 
fairly, except that the mute has to be pronounced in the same syllable as 
the h (she-pherd etc.). In Laconian however 6 was apparently our th (in 
thin) at an early date, as is shown by the frequent spelling 0, which may 
represent a further development. In other dialects likewise the tendency 
to make @ a spirant seems to have worked fairly early. In the Hellenistic 
period we may probably assume that the spirant pronunciation (p=f, 
6=th, y=Scotch or German ch) was fully developed in most parts of the 
Greek-speaking world, though it may have lagged in some. After a, or 
another aspirate, the spirant never developed: φθάνω is in MGr ftdno, 
αἰσθάνομαι is esténome. This only partially applies to ¢. The MGr 
spirant pronunciation of «x and wm before r (€mra=eftd, ὀκτώ =ochté), 
shows its earliest indications in the period to which our uncial MSS 
belong. 

The Liquids X and p, and the nasal liquids y 
(before x, y, x, €), μ᾿ and ν, being practically identical 
with the modern sounds, need no detailed description. As in nearly 
all European languages except (southern) English, the p is trilled: when 
initial, or following an aspirate or another p, it was breathed (Welsh rh), 
but in our period the breath in initial p was feeble. 

Sibilants The Sibilant o was always breathed (our hiss), 
ἦ though the voiced sound (as his) was heard before voiced 
consonants, as in σβέσαι, Σμύρνα. For this z sound ¢ was often written. 
This symbol, which in the classical period seems to represent sometimes 
dz, sometimes 2d, was already tending towards the simple z, as in MGr. 
The pronunciation of the closely related combinations oo and rr is very 
difficult to determine. There is some reason for assigning the value ts 
to the former, and tth (as that thing pronounced in one word) to the 
latter ; but this is conjectural. The combinations ἔ and Ψ are always 
breathed. 


Liquids, 


There remain the Breathings,’ and‘. The former, 
which the Hebrew alphabet represented by & (initial) 
is heard at the beginning of every English word with vowel initial in 
normal pronunciation : it is the glottal catch which, except in singing, 
we can hardly avoid sounding before the vowel. The corresponding 
breathed sound, our h, was already obsolete in some of the Greek dialects 
of classical times, and must have been faintly heard in many districts in 
the Hellenistic period, apart from those which inherited the early pszlosis. 
In Palestine we may fairly assume that the aspirate survived intact, as 
was natural in a country where the native speech retained this element 
so clearly. A native name like »55m was represented by ᾿Αλφαῖος : the 
Vulgate transliteration Alphaeus illustrates the enfeeblement of h in 


Breathings. 


46 A GRAMMAR OF NEW TESTAMENT GREEK. [ὃ 25-26 


other regions and at a later time. The fact that effects of initial aspirate 
like καθ᾽ ἡμέραν not only survived but even produced analogic forms like 
καθ᾽ ἔτος (see § 40), is evidence as far as it goes that the complete psilosis 
of MGr was only local in the earlier centuries of Hellenistic. 


The above resumé of results will be enough to show how 
complex the subject is. A history of Greek pronunciation in 
the Hellenistic period is greatly needed, showing both when 
and where the various developments first appeared which 
issue in the system, or rather systems, of MGr. Such a 
history would have an important bearing on textual questions. 
The pronunciation of Greek in Palestine can be ascertained 
with considerable accuracy from the very numerous Greek 
words borrowed by Aramaic and Aramaic words transliterated 
into Greek. Since, however, the bulk of the NT was written 
outside Palestine, there seems no reason for taking up one 
part of a complicated investigation, to complete which would 
be impossible within our limits. The sections which follow 
on Orthography will raise a number of points bearing on 
pronunciation, supplying some of the evidence for summary 
statements already made, and discussing the relations of these 
facts to the phenomena of our MS tradition. 


MopERN PRINTED GREEK. 
Punctuation. 


§ 26. Ancient writing knew very little 
of so obvious a help to reading as punctua- 
tion. As early as iv/B.c. we find the mark : in the 
“ Artemisia” papyrus at the end of some sentences. The 
great grammarian Aristophanes (ili/B.c.) invented the stops 
(-) and (.), to which he gave the reverse values to those for 
which we use them; also a comma (-), placed half-way 
between the position of the colon and the full stop. Literary 
papyri, especially those of lyric poets, who needed such helps 
to reading more than others, show these punctuation marks 
earliest. Thus the Bacchylides papyrus (i/B.c.) shows the 
colon freely at the end of sentences, and “ generally, it would 
seem, correctly,” says Kenyon (Bacchylides, xxi). See further 
E. Maunde Thompson Palaeography, 60. 


Punctuation ; 


§ 26] SOUNDS AND WRITING. 47 


The oldest NT uncials have none of these 
adjuncts. “In 8B the first hand very rarely 
shows any points, and the words are not 
divided except sometimes when a change of subject brings 
in δὲ a new line or in B a small blank. Later we find a 
single dot, as in the second hand of xB and the first in 
ACIN*®PQZE ; sometimes as in A a comma and a double dot. 
In NR 0115 the dot is placed at the top, the middle or the 
bottom of the letter without distinction. The dot is most 
freely used in FP and G?.” So Gregory (Ti® ΠῚ. 111), who 
gives a specimen from F in 1 Tim 3", with each word 
divided from the next by a dot. He goes on to observe that 
obviously no argument towards a right punctuation can be 
drawn from the barrenness of the earlier or the abundance 
of the later signs. 


in Earlier 
Uncials, 


A fuller system is observed by the later 
uncials, including ELMP?, and they are here 
as in some other features anticipated by W: 
though its punctuation is scanty, it goes decidedly beyond 
xB. The single dot occurs on the average less than three 
times per page, being most frequent in the first part of Lk, 
and least so in Mk. “The double dot (:) occurs 12 times 
in Mt, 6 in Jn (excluding the first quire), 23 in Lk, and 
11 in Mk (7 are in the first four chapters).” Sanders (The 
Washington Manuscript 12) goes on to observe that (:) was 
“a decidedly strong punctuation,” used generally at the ends 
of paragraphs. There is also “a substitute for punctuation 
formed by leaving small blank spaces between the phrases. 
These occur frequently and regularly in all parts of the 
MS,” and are said to correspond fairly well with the number 
of στίχοι (see below), and more closely still to the ῥήματα. 
“They doubtless coincided with the sense-divisions used in 
reading.” Sanders compares these phrase-lengths in W, and 
the στίχοι in D, with the introduction of each phrase by a 
capital in 4, so as to deduce “an ancient system of phrasing, 
used in reading the Scriptures in church service.” It may 
have originated as early as ii/A.D. 

Paragraphs are found even in the earliest 
uncials: an account of those in W, as com- 
pared with x, B and D, may be seen in Sanders 15 ff. On 


in Later 
Uncials. 


Paragraphs. 


48 A GRAMMAR OF NEW TESTAMENT GREEK. _ [ὃ 26 


the στίχοι, so conspicuous in D, where the text looks like 
poetry in very short uneven lines, and ῥήματα or ῥήσεις 
“sentences,” a very similar but independent division, see 
J. R. Harris Origin of Ferrar-Group (Cambridge, 1893), 8 ff. 

It will be clear that there is little probability that any 
punctuation worth counting such was present in the NT 
autographs. If they had any of Aristophanes’ points—never 
found in papyri dated B.c.—or those of different systems 
which are known to have existed in Aristotle’s time,! we have 
no proof that they underlie the meagre punctuation of W and 
later MSS. 


Modern 
Punctuation. 


Passing to our modern system, we should 
begin with a full recognition that it is purely 
modern. Apart from the use of (;) for the 
query, and (°) for colon and semicolon alike, the punctuation 
of our printed Greek Testaments is on exactly the same 
footing as that of their English versions. It is simply in 
essence a form of commentary; and the modern editor is on 
every page compelled to choose between alternative punctua- 
tions, involving different interpretation, where the only 
ancient authority is that of patristic comments or early 
versions. Since these go back to periods considerably ante- 
dating our best uncials, they have naturally the weight in 
many cases of a primitive tradition, which no wise exegete 
would ignore. But as little would he consent to be bound 
hand and foot by interpretations: which do not depend on the 
autographs, and may be no more than guesses by readers who 
were not by any means better qualified from all sides than 
ourselves. When therefore we use an extremely careful 
edition like that of WH, where punctuations in text and in 
margin are constantly determining the meaning for us, we 
must always be careful to realise our freedom to take our 
own line on sufficient reason. Rarely—as when in Jn 13: W 
punctuates ἐγένετο οὐδὲ ev" ὁ γέγονεν ev | avtw fon —their 
interpretation may have behind it the punctuation of the 
oldest MSS in which such marks appear at all. But even 


1 On these see Kiihner-Blass i. 351-3, Mayser 48-50. Mayser notes that an 
instance of the παράγραφος, a stroke which points to the end of a sentence, is 
found in v/B.c. in a Laconian inscr. More rarely still in papyri appears the 
double point, which may also be seen on inscrr. 


§ 26-27] SOUNDS AND WRITING. 49 


there, if exegetes insist on the reading of AV and RV, we 
cannot oppose them successfully on the authority of W: our 
arguments must be exegetical, and the traditional punctua- 
tion seen in W will count for little. 

It is not the function of a Greek Grammar under these 
circumstances to lay down principles for punctuation. Our 
system being purely modern, we insert commas just where we 
should insert them in a corresponding sentence of English. 
Since any kind of editing in modern form involves in many 
places the editor’s decision between rival interpretations, it is 
well frankly to recognise this, and insert these helps to 
reading freely. At the best they will not be equal to those 
we employ. Quotation marks—which many uncials use for 
OT citations—might be multiplied to advantage. There is 
one further difference between ancient and modern writing 
which we might well reduce. The absence of facility for 
indicating parentheses introduces complications into our 
understanding of a great many passages. Often we should 
simplify a passage considerably by taking out a parenthesis 
and putting it underneath as a footnote. Thus in Jn 49 
the last sentence is taken as the author's comment in RV, 
and in a modern work would naturally become a footnote: 
still more clearly Ac 1155, which interrupts the speech with 
matter extraneous to it. Further instances may be Heb 34, 
1 Co 72 ἀλλ᾿ εὖ kal . . . χρῆσαι, ib. 1577, Tlustrations, of 
course, could be multiplied indefinitely, and those given 
already do not pretend to be indisputable.? 


§ 27. Very few of our modern accessories 
existed in the early periods of the NT, or 
appear in our oldest MSS. Sreathings begin 
to appear in v/A.D., as we shall see in ὃ 40. Accents had 
been invented long before, but seem only to have been used 
in poetical texts, such as the Bacchylides papyrus (i/B.c.): 
see § 36. It was a long time before MSS even began to 
divide words—a point in which Greek was curiously behind 
Latin, where words are separated in early inscrr. The use of 


Other modern 
accessories. 


1 An attempt is made in The Modern Reader's Bible, by R. G. Moulton 
(New York and London, 1907), to present the English RV text with these 
accessories of modern printing. 


VOL. Il1.—4 


50 A GRAMMAR OF NEW TESTAMENT GREEK. [ὃ 27 


“ {ota subscript,” convenient as distinguishing sundry flexions, 
but not answering to any living feature of speech, was dis- 
continued some time before the NT began to be written, and 
only returned with Byzantine scholarship: indeed in the 
strict sense it is a practice of xii/A.D. and later, for not 
till then was the unpronounced ¢ written underneath. To 
complete the difference between a modern printed Greek 
Bible and an early MS thereof, we have the use of the small 
and convenient “minuscule” script, in place of the clumsy 
and space-filling “ capital” letters, which though in early use 
for non-literary purposes, only in ix/A.D. began to be applied 
to books. Further details on most of these subjects will be 
given below. 

One of these accessories, however, stands 
on a different footing, the Diaeresis, which 
figures largely in our oldest MSS and in the papyri. It is 
used to distinguish vowels, especially ὁ and v, which begin a 
syllable: thus we find ina, ὕπο ete. Our use of it is a special 
application of the same principle—if we may call it a 
principle which is applied only in a minute proportion of the 
instances where on the above definition it should appear. 
We employ it to distinguish vowels which are to be pronounced 
separately, and keep them from being merged in the vowel 
before. Thus TAIOC might be read Γαῖος (as WH actually 
print), had we not evidence that the Latin Gaius was trisyllabic, 
so that we should write Γάϊος : in this case the accentual 
difference makes the diaeresis less necessary. In MSS the 
diaeresis often preserves evidence of pronunciation, as when 
N writes ηἰὐὐλοκηοὰ in Mt 317, even dividing the line after 
the first element in the diphthong. The diaeresis often 
affects transliterations of foreign words. Thus in Mt 115] = 
Lk 10% D has yopofaiv and βεθσαειδα (Bedcaida in 
Lk), implying by two typical devices (diaeresis and aee for 
at) the tetrasyllabic pronunciation of both names. On the 
other hand Jn 115: ᾿Εφρέμ xL reinforces the absence of 
diaeresis in the uncials which write “E¢paiu. Our uniform 
accent system spares us the necessity of using the diaeresis 
very often: πρόιμος, πρωΐ, πρωινός, Πτολεμαίδα, Awis, ete. 
might as well be left without, while "Ayaia, “Pwpaiorti and 
the like might be ambiguous. 


Diaeresis. 


§ 28-29] SOUNDS AND WRITING. 51 


SOUNDS AND ORTHOGRAPHY. 


§ 28. Educated Greek writing in our 
period had practically lost the earlier phonetic 
character. Attic literature dominated the 
orthographic tradition, though Hellenistic pronunciation 
diverged widely from Attic. We have accordingly a great 
difference between the conventional spelling and that of less 
educated people, who tended in various points to write as they 
and others spoke. The great NT uncials present a consider- 
able number of spellings thus depending on the pronunciation 
current in the period when they or their originals were 
written. How far we ought to follow them in their devia- 
tions from conventional spelling is an open question. 
There are some points in which the evidence of papyri and 
inscriptions shows that a particular spelling was widely 
current in i/A.D., and may reasonably be traced to the author. 
This is made especially plausible in the not infrequent cases 
where the said spelling was extinct before the actual period 
of the seribe. This point is well brought out for Codex 
Bezae by Rudberg, Nétlicher Text u. Nomina Sacra (Upsala, 
1915). He notes that while D perpetually confuses az and 
e, εἰ and z, it hardly ever confuses ἢ and ¢, and never 
(according to von Soden) 7 and εὖ, although in v/vi A.D. 4, ἢ 
and ev were indistinguishable: he argues that we have here 
a proof that the orthography of D is that of u/ap. He 
proceeds on the same lines with sundry other orthographical 
peculiarities of the Bezan text. 

The notes which follow are restricted to matters of 
spelling, and the questions of pronunciation affected. Where 
flexions are concerned, reference must be made to the 
Accidence. 


Spelling no longer 
Phonetic. 


Accentuation, 


§ 29. The accents with which Greek has 
been written since the Hellenistic age are the 
invention of the great grammarians who 
tried to preserve a record of the classical language when it 
was in danger of obscuration. In their time the character 
of the accent was changing from pitch to stress. The MGr 


Accents, ancient 
and modern. 


52 A GRAMMAR OF NEW TESTAMENT GREEK. [8 29 


accent, which remains with few exceptions on the same 
syllable as in the ancient language, is just like our own; and, 
as in English, the stress affects the quality of all syllables, 
stressed or unstressed. Thus ἄθρωπος man has stress on the 
first syllable, and the second and third syllables in consequence 
have the same vowel: ὦ becomes long again in the plural, where 
the accent falls on the penult. In classical Greek there was 
a “musical” accent, the tone involving a higher note but no 
sort of stress. We have this musical accent in English, and 
it plays a very important part. But it is perfectly free, 
depending on the shade of meaning intended by a speaker, 
and differing very much with different individual speakers: 
in Greek the tone was tied to the word or word-group, and 
was capable of no variation. It was a fixed element, almost 
as much as a similar but more elaborated tone-system is in 
Chinese. We recall the well-known story of the actor 
Hegelochus, who in declaiming a line of Euripides ending 
with γαλήν ὁρῶ (“1 see a calm”) pronounced a circumflex 
instead of an acute, and sent the audience into roars of 
laughter: γαλῆν ὁρῶ = “I see a weasel.” 

We need not pause to state in detail the 
rules of the Greek accents, which are fully 
given in the ordinary Greek grammars. The acute accent 
marks the rising inflexion of the voice: the second syllable 
of our interjection “ Really?” (expressing surprise or in- 
credulity) usually shows this intonation. The falling tone 
(heard in the same syllable of “ Really!”) is written only 
when an acute falls on the last syllable of 
a word and is changed by rule to “grave”: 
this happens when the word as it comes in a sentence does 
not precede an enclitic (see below), or a stop (comma, colon, 
full stop or query). The interrogative tis however keeps its 
acute accent under all conditions. The falling inflexion 
belongs naturally to all syllables which are not marked with 
accents. In addition to the acute (’) and the grave (ἢ) there 
was the circumflex accent, which denoted the 
combination of the two (~="): it was con- 
fined to long syllables, where the voice rose in pitch during 
the first half (technically mora) and fell in the second. We 
may represent the three accents in musical notation, premising 


Acute. 


Grave. 


Circumfiex. 


§ 29] SOUNDS AND WRITING. 53 


that the notes would vary in pitch with different individuals, 
and that the intervals would not be constant :— 


ee ae 
———— 
eee ee ee ra er τ 1} τ τ Ξ-" 

φέ- ρὼ πῶς 


(The time of a mora is represented here by ἃ crotchet.) 


The following rules are a summary of the 
general principles governing Greek accentua- 
ion :— 


Rules of 
Accentuation. 


(1) The limits of the position of an accent depend on the “ three- 
syllable law,” by which the rising inflexion cannot stand farther back 
than on the third syllable from the end of a word. Only in the case of 
a word ending with a trochee (7) can the equivalent of three short 
syllables at the end of a word stand together unaccented. 

Thus (a) if a word ends with a short syllable, the acute may fall on 

the ultima, as λεκτός, an “ oxytone” word, or on the penult, as 
παρθένος, a “paroxytone,” or on the antepenult, as λέγομεν, a 
“»roparoxytone” ; the circumflex may fall on the ultima, as τιμῶ, 
a “‘perispomenon” word, or on the penult, as τιμῶμεν, a “ pro- 
pertspomenon.” (N.B.—If a word ends with ~ and the penult 
has an accent, it must be a circumflex.) 
(6) If a word ends with a long syllable, the acute may fall on the 
ultima, as τιμή (oxytone), or on the penult, as τιμώντων (par- 
oxytone) ; the cercwmflex can fall only on the ultima, as τιμῶν 
(perispomenon). 


The rationale of these restrictions appears as soon as the circumflex is 
resolved into acute + grave, and each mora—short syllable or half of a 
long one—has its accent supplied. Thus τίμωντὼν and τιμῶντὼν are seen 
to be impossible, since in each case the rising inflexion is followed by 
more than two morae with falling tone. Τιμώντων, accented in full, would 
be τὶμώντων, and is according to the rule. 

(2) For purposes of accent, syllables are not long unless the vowel in 
them is a long vowel or diphthong. Thus in λαϊλᾶψ᾽ the second syllable 
is short, and the circumflex falls on the first syllable accordingly, 
although the word is a spondee for metrical purposes. Greek grammarians 
tell us that we must write κῆρυξ and φοῖνιξ : if this is correct, we must 
assume that the v and ¢ were shortened in the nominative before &. 

Final -ac and -o. (not followed by a consonant) had only one mora 
each, except when they were locative terminations or optatives, and of 
course when they were contracted. Thus we have οἶκοι, χῶραι nom. pl., 
but οἴκοι, ᾿Ισθμοῖ, loc. sing. ; τίμησαι imper mid., τιμῆσαι infin. act., but 
τιμήσαι opt. act. 


54 A GRAMMAR OF NEW TESTAMENT GREEK. [8.99 


(3) Since accents were in existence long before the contraction of 
vowels or other later modifications which arose in the historical period, 
the accentuation must follow the earlier conditions. Thus ripdoper 
(i.e. τὶμάὸμὲν) became τιμῶμεν, ἑσταώς (ἑστὰώς) ἑστώς, πόληος became 
πόλεως When the quantity of its last two vowels was inverted. When ra 
ἄλλα is contracted into one word we write τἄλλα, not τἄλλα, since the crasis 
joins grave + acute, and the rising inflexion comes on the second mora. 

(4) Against these orthotone (7.e. accented) words may be set two classes 
of words which were without an accent under specific conditions. 
Proclitics linked themselves to the word following, 
and took an accent only when that word threw its 
accent upon them. Thus ἐκ τούτου, ἔκβηθι, οὐκ ἔστι, οὔποτε. Enclitics 
(except when standing at the beginning of a sentence) threw their accent 
back upon the last syllable of the preceding word, 
unless this would involve two acutes following. Thus 
τινὲς μὲν. . . (Phil 17°), ἔκ τινος, εἴ τινων, but ἄλλα τινά : ὁποῖά τινα, but 
ὁποίων τινῶν. Proclitics include the article forms ὁ, ἡ, οἱ, ai, the prepositions 
εἰς, ἐξ (ἐκ), ev, the conjunctions εἰ and ὡς, and the negative οὐ (οὐκ, οὐχ). 
But οὐ takes accent (οὔ) at end of sentence, or standing alone. [ἢ reality 
all prepositions are proclitic, for (¢.g.) ἀπὸ has no rising inflexion, and its 
own accent is paroxytone when standing free or after its case. Enclitics 
include the pronoun forms μοῦ, pol, μέ : σοῦ, σοί, σὲ (except when emphatic) : 
ris (indefinite) and the indefinite adverbs πού, ποτέ, πώ, mas: the 
particles γέ, τέ: and the present indicative of εἰμί (except 2 sg. εἶ), and 
φημί, φησί. (Ἔστί at the beginning of a sentence becomes ἔστι : so after 
οὐκ, μή, εἰ, ὡς, καί, ἀλλά, and τοῦτο, and when it means “exists” or “ is 
possible.”) 

A proparoxytone or properispomenon or a proclitic, followed by an 
enclitic, receives an acute accent on its ultima ; an oxytone keeps its acute 
without change to grave. A perispomenon is unchanged, as is a par- 
oxytone, but if the enclitic has two syllables, the second is accented. If 
a series of enclitics follow one another, each throws an accent back on the 
preceding, as εἴ ris τί σοί φησιν. But pov, σοῦ throw an acute upon the 
preceding word and receive it from a following enclitic, e.g. σύνδουλός 
σού εἰμι (Rev 19! 22%), The rules for enclitics account for some 
combinations which seem to break the ordinary principles given above : 
thus ᾧτινι, ὥστε. 

(5) The accent of nouns, adjectives and pronouns must generally be 
learnt from observation, and there are few general rules. Monosyllabic 
3rd decl. nouns usually are oxytone in gen. dat. sing. 
and dat. pl., perispomenon in gen. pl. Other nouns 
keep the accent on the same syllable as in nom. sing. unless it is 
forced forward by the rules of accent, as θάλασσα gen. θαλάσσης. The 
gen. pl. in 180 decl. nouns (not adj. or pron.) is perispomenon (since -ῶν 
comes from -έων, and that from (-ήων), a-(c)wv). Vocatives sometimes 
retain their primitive recessive accent (1.6. accent as far as possible from 
the end of the word): thus πάτερ from πατήρ, ἄδελφε from ἀδελφός. 
(This arises from the original enclitic character of vocatives when not 


Proclitics. 


Enclitics. 


Noun Accent. 


§ 29-30] SOUNDS AND WRITING. 55 


opening a clause.) It may be noted that (as in MGr) common nouns 
when taken to serve as proper names often suffered a change of accent. 
(6) Finite verb forms are recessively accented—a consequence of their 
primitive enclitic condition. Infinitives and participles, being nouns or 
adjectives, do not come under this rule. In compound 
Verb Accent verbs the accent cannot go behind the augment or 
reduplication. Apparent exceptions to the recessive rule are historically 
due to contraction: there is however the peculiar case of strong aorist 
imperatives, εἰπέ, εἰπόν, λαβέ, λαβοῦ, which (like the case of ἔστι above) 
are survivals of the primitive condition retaining the verb’s original accent 
when standing first in a sentence, as imperatives naturally did. 
(7) In a few words which suffer crasis—the fusing of two words into 
; one by contraction—the accent of the first word is 
Crasis. ignored. See § 32. 


§ 30. The record of classical accentua- 
tion is in many respects seriously imperfect, 
and it,is probable that our modern printed 
texts differ not infrequently from the genuine pronunciation 
of their authors. This is well seen when we study classical 
texts preserved for us in ancient papyri. Professor Wilamo- 
witz goes so far as to say (Sappho und Simonides*, 100 f., ap. 
Sonnenschein in Zhe Year's Work in Classical Studies, 1913, 
p. 102) :— 


Imperfection of 
accent record. 


We now possess so many remains of ancient books that we see clearly 
how late, how rare, and how incomplete is the indication of accents, and 
how far it departs from that which is customary among us at the present 
day. Noone can any longer dispute the fact that our accentuation is a 
product of the time of Photius. 


The statement is not intended to suggest that we must simply 
acquiesce in ignorance, for Wilamowitz declares that we are 
“well able to go behind the Byzantine period,” and are bound, 
to do so. It must not be forgotten that the MGr accent, 
scientifically used, can help us a great deal in verifying our 
accentuation of ancient texts, goimg back as it does to the 
period of the Κοινή. 
: Our specific information for the accentua- 
Accentuation of |. i 
NT text. tion of the NT text comes necessarily from 
later authorities. Sir E. Maunde Thompson 
states + that accentuation, only occasional even in literary texts 


1 Introduction to Greek and Latin Palaeography (Oxford, 1912), 61 f. 


56 A GRAMMAR OF NEW TESTAMENT GREEK. [8 30 


of the papyrus period, begins to become general in iii/A.D. 
The oldest literary texts showing accents at all regularly are 
the Bacchylides and the Aleman, both i/B.c.; Sir F. 6. 
Kenyon observes that lyric poets seem to have needed 
helps to the reader more than other texts. The earlier 
uncial MSS on vellum have no accents at all: we have to 
wait till vii/A.p.1. There are a few instances of the circum- 
flex in D, perhaps in N, and in 4 and F,G;, but mostly 
alien from our modern use. The later uncials are accented 
on our system generally, and late correctors have equipped 
B and C; but the best of them are least accurate in this 
respect. Gregory’s table (Prol. 100 ff.) may be consulted 
for the evidence of these MSS as to the accent of doubtful 
words. 

Soon after the date A.D.—a period when 
the Κοινή began its first new period—the old 
musical accent developed into a pure stress ; 
and we may assume that the NT documents were from the first 
pronounced with the accentual conditions familiarin MGr. Τί 
we read the words aloud with a stress upon the syllables 
written with an accent—all three accents being now equiva- 
lent—we shall be practically compelled to reduce to a 
minimum the difference between long and short vowels, 
imparting the quality of length to the stressed syllable alone 
(cf. Prol. 54 υ.5). This is practically the pronunciation of the 
modern language. Quantitative levelling was not complete 
in i/A.D., but the distinction between o and ao, 4 and a, was 
becoming very slight. It follows that when texts were 
transmitted to any extent orally, such distinctions as that 
of ἔχομεν and ἔχωμεν were very easily lost. The new 
stress accent began to affect the forms of words. The 
Hellenistic νοσσίον for νεοσσ. has been attributed to this 
cause; Thumb rejects this (Brugmann Gr. 76). A clearer 
case is that of Kretschmer’s Law, by which an unstressed 
vowel after a liquid or nasal dropped out when the same 
vowel stood in the previous syllable. Thus Bepvixn answers 
to an older form Βερενίκη: see Brugmann G7.‘ 80, and 
§ 33 below. 


Pitch accent 
becoming stress. 


1 Gregory Prol. 99. 


§ 30] SOUNDS AND WRITING. 57 


Some points may be collected in which the accent- 
uation of NT words is doubtful, or specially signi- 
ficant :— 


Specific accent 
questions in NT. 


(1) There are some words in which we have evidence of a change of 

pronunciation between classical and Hellenistic times :— 

(a) Certain 2 aor. imper. forms—ié¢, εἰπέ, λαβέ, ebpé—which in Attic 
retained a primitive oxytone, have in Hellenistic succumbed to 
levelling. (Moeris, s.v. ἰδέ, p. 193.) 

(6) Shortening of quantity took place in the penult in many nouns in 
-wa. Thus in Hellenistic we find ἀνάθεμα (in poetry ἄνθεμα) in 
place of ἀνάθημα, and even words like σύστεμα, not *-craya, show- 
ing that the shortening is late. The wide extent of this pheno- 
menon, which is probably due to the analogy of nouns in -σις, 
makes κλίμα, κρίμα (80 MSS), μίγμα preferable to the class. pro- 
perispomena. In the case of χρίσμα we have definite evidence 
that the. was short: see W. M. Lindsay The Latin Language, 
30. Possibly the case of κῆρυξ and φοῖνιξ may come here: for 
the grammarians who prescribe this accentuation see Chandler’s 
reff., § 669. The ultimate shortening of . and v before € may 
perhaps have been Hellenistic, in which case we might accentuate 
Φῆλιξ, κῆρυξ and κηρύξαι! in Hellenistic, leaving the question 
open whether κήρυξ and κηρῦξαι should be retained for classical 
times. See Lobeck Paral. 411. In his Phryn. 107, Lobeck 
mentions a number of dissyllabic nouns which shortened their 
penult in later Greek. We may add ψύχος. Shortenings like 
θλίψις (30 MSS) may be defended on the same lines, but we 
should need special evidence to justify ῥίψαν (as several MSS), 
συντετρίφθαι. Σπῖλος is a mere mistake, for the. is short from 
its first appearance, which is not early. See in general Lipsius, 
Gr. Unt. 31-46, summarised by W. F. Moulton, WM 57 n. 

(c) A different class of shortening may arise in the nom. ace. sing. of 
nouns in -eva where confusion with those in (e)ia is possible: the 
complete identity of εἰ and. in popular speech helps the confusion 
by bringing -eia and -ia nouns together. Ἐριθεία and ἀρεσκεία 
are claimed for the -a class on the strength of derivation from 
verbs in -evw. See on these nouns below under Word-formation. 

(2) In the following NT words the accent is questioned, or accentua- 

tion distinguishes different words or forms :— 

ἀγόραιος and ἀγοραῖος are differently distinguished by grammarians. 
Zonaras has ἀγοραῖοι for οἱ ἐν ἀγορᾷ ἀναστρεφύμενοι ἄνθρωποι, and 
ἀγόραιος aS ἡ ἡμέρα ἐν 7) ἡ ἀγορὰ τελεῖται. If that is correct, we 
must write ἀγόραιοι in Ac 1988, and make ἀγοραῖοι the nom. of 
the noun in 175; but Ammonius, who is eight centuries older 
than Zonaras, gives an entirely different distinction. HLP have 
ἀγοραῖοι in Ac 1988, 


180 FHKMUL4 in Lk 4”, and HL 61 in Ac 10: the MSS for κήρυξ and 
κηρῦξαι seem to be few (Gregory Prol. 101). 


58 A GRAMMAR OF NEW TESTAMENT GREEK. [§ 30 


ἄγων from ἄγω : ἀγών a noun. 

ἀδελφός has the old voc. ἄδελφε, as seems proved by the survival of 
ἄδεφλε in Pontic MGr (beside ἀδερφέ). 

ai from art. : ai relative. 

ἄλλα is neut. pl. of ἄλλος : ἀλλά but is the same become proclitic for a 
conjunction. 

ave=Up: ave 2 aor. subj. of ἀνίημι. 

ἀπόδεκτος, as a compound verbal assumed to be of two terminations, 
will retract accent : see however Kiihner-Blass i. 538, WS 69. 

apa=curse : ἄρα then, apa asks question. 

ἀχρεῖος said by gramm. to be non-Attic for ἄχρειος. 

γλωσσοκόμον is defended by Schmiedel (WS 140), on the ground 
that the second element is active : WH print γλωσσόκομον. 

δεινά terrible things : δεῖνα so-and-so. 

διά through : Δία acc. of Ζεύς. 

εἰ if: et thow art, or in εἶ μήν (also written εἰ μήν) verily. 

εἰπόν imper. : εἶπον indic. The grammarian Charax tells us that the 
former was Syracusan Greek ; the latter appears to be Attic 
(Lobeck Phryn. 348). Which belongs to the Κοινή is not very 
certain, but editors in Ac 3836 print εἰπόν (so B’—contra LHL, 
1, 61). 

ἐκλεκτός has three terminations and is oxytone: Kwhner-Blass i. 538, 
Chandler 199. 

ἔρημος, ἕτοιμος, ὅμοιος are said by gramm. to be Attic, while properisp. 
is Homeric or Ionic. Supposing this true, it would not be proof 
that we should not write ἐρῆμος etc. in NT. But MGr has 
ἔρημος, ἕτοιμος, ὅμοιος, which is better evidence. 

ἔχθρα enmity: ἐχθρά fem. of ἐχθρός hostele. 

ἡ def. art. : 7 rel. 

ἴσος is doubtless the Κοινή form (Attic), though the Epic ἶσος occurs 
in late poetry. 

καθαίρω purge: καθαιρώ pull down. 

κλείς key : κλεῖς keys. 

μωρός is generally preferred, but gramm. give μῶρος also as Attic. 
MGr seems to speak for μωρός, but an inter). μῶρε survives in 
Pontus : there was presumably dialectic variation. 

Νύμφαν Nympha (acc. fem,) : Νυμφᾶν Nymphas (masc.). 

‘6, oi nom. masc, sing. and pl. def. art.: 6, of neut. sing. and masc, pl. 
relative. 

ὅμοιος 868 8.0. ἔρημος. 

ὄργυια and ὀργυιᾶ alternate in our authorities. 

πότε, ποῦ, πῶς interrog. : ποτέ, που, πως indef. 

πρωτότοκος first-born: πρωτοτόκος first-bearing. See on this general 
distinction under Word-formation. 

σκῦλον and στῦλος should be left unchanged : see WS 68. 

τεσσαρακονταέτης (and the like) of time :— -erjs elsewhere—thus in 


Ac 723 1318 (so 81) parox., but ἑκατονταετής Rom 4198 (not DeL). 
But the evidence is insufficient : see Lobeck Phryn. 406. 


§ 30] SOUNDS AND WRITING. 59 


ris ri etc. interrog. : τις ru ete. indef. 

τροχός wheel: τρόχος cowrse has no real claim in Jas 3°. 

φάγος is printed as a paroxytone, on the ground that it is a substan- 

tive, the adjective being φαγός. MGr dayas, fem. φαγοῦ, 
“‘courmand,” are some evidence against this. 

(3) When oxytones or paroxytones were adapted as proper names, 
the accent was drawn back. This is perhaps due to the strong influence 
of the vocative in personal names, and the survival of 
the old rule by which the vocative took recessive accent : 
cf. πάτερ, ἄδελφε, and note how the predominance of 
the vocative changed the accent of μήτηρ (Skt mata, with same accent 
evidenced by O.E. moder). In NT we have Τύχικος, Φίλητος, ᾿Επαίνετος, 
"Epaotos, Πύρρος, Βλάστος, Kapros, ᾿Ονησίφορος (for parox.), Σωσθένης, 
‘Eppoyévns, Διοτρέφης : cf. WS 70. The name Χριστός was not changed, 
since it was never in the first age treated as areal proper name. ᾿Αχαϊκός, 
not having been a common noun to start with, was not altered. The 
rule does not seem to be conditioned by convenience of differentiation, 
for names like Ὑμέναιος, Τρόφιμος, ᾿Ασύγκριτος retain their accent un- 
changed. The principle survives in MGr: see Thumb Handbook, 
§ 38. 1. 


(4) Oxytone adjectives retract the accent when they become nouns: 
thus κρύπτη from κρυπτός. This is also MGr (Thumb /.c.). 


Recession of 
accent. 


(5) Latin words were sometimes accented in a manner which re- 
minds us that Latin and Greek accentuation differed altogether in 
quality. (Thus we get the place-name St. Heléna ulti- 
mately from the Greek “Ἑλένη, but the personal name 
Helen from Latin Hélena, where the accent was changed 
by the short penult.) The Latin Ohristidnus became Χριστιανός, as it is 
still in MGr: similarly other adj. in -a@nws. Words in -tnus, but not those 
in -énus or -anus, retain accent on penult, as ᾿Αλεξανδρῖνος, λιβερτῖνος. The 
analogy of genuine Greek words in -avés, -nvés presumably worked here. 
Generally however we accent Latin words in accordance with their 
original form: thus Λίνος, Tiros, Πειλᾶτος, Τάιος (trisyll.), Kovdpros, 
Σεκοῦνδος, Πρισκίλλα, ete. 


Accent of 
Latin words, 


(6) The current accentuation of Semitic words is, as Schmiedel notes 
(WS 76), often governed by pure caprice. The only intelligible principle 
would be to set the accent always on the tone-syllable 
of the original Semitic. Our difficulties in that case 
would arise only where words had been partially 
Hellenised ; and even here, unless popular etymology had produced a 
really new form, or the word could be shown to have taken an indepen- 
dent place in Greek—as in our treatment of the place-names Paris and 
Vienna—it would be safe to keep the tone-syllable unaltered if possible. 
When Hebrew names were inflected in Greek declension the variations 
of case naturally affected the accent in places. It is needless to discuss 
the application of these principles, as the number of words affected is 
extremely large. 


and of 
Semitic. 


60 A GRAMMAR OF NEW TESTAMENT GREEK. [§ 30-31 


(7) WS (71) gives a selection of passages in which (except in those 
marked*) a different accentuation would involve a change of sense. 
The forms are quoted here as they stand in WH, the 
alternative, whether probable or not, being added in 
brackets :—pevei 1 Co 8318 (μένει), κρινοῦσιν 6? (WH 
mg. κρίνουσιν), ἴαται Mk 539 (tara.—as Ac 953), καθῆσθε Lk 2230 (κάθησθε 
indic.), βαθέα Rev 23 (Badea! an uncontr. pl. from βάθος), ἅγια Heb 9? 
(ayia), ἢ before πόρνος 1 Co 51! (ἢ), ὅμως 147 (ὁμῶς -- ὁμοίως), φώτων Jas 117 
(φωτῶν men!), τίνες Heb 316 (τινές, as in AV), τίσιν 317: (τισίν), τῷ 1 Co 
158 1616, 1 Th 456, Mt 244! (τῳ, Attic for τινί, and obsolete), ἔσυνίων 
Rom 311 (συνιῶν---8566 8 86n 1 (8)), ἔσυνίουσιν Mt 1318 (συνιοῦσιν), ἄρα 
Gal 217 (dpa), οὐκοῦν Jn 1887 (οὔκουν), αὕτη Mt 2259 (WH mg. αὐτῇ), Lk 2? 
(αὐτή), Rom 710 (αὐτή), 1 Co 7}? (αὐτή), αὐτή Lk 287 713 847, Rom 16? (αὕτη), 
εἰμί Jn 754 36 148 1724 (εἶμι !—obsolete), EAa@v Lk 1959 2197 (Ἐλαιών--- 
see § 61 (b), Prol. 49, 69, 235), μακρά Mk 1240 (μακρᾷ), ΠΟΡΝΩΝ Rev 175 
(indeterminate between πόρνων m. and πορνῶν f.), κερδανῶ 1 Co 9?! 
(κερδάνω subj.—see § 95), ἔπίμπρασθαι Ac 28° (πιμπρᾶσθαι---866 § 95), 
ἐκφύῃ act. Mt 2492, Mk 1378 (ἐκφυῇ pass.—see ὃ 95), *amoxvet Jas 115 
(ἀποκύει--ὃ 95), μένει Jn 1417 (μενεῖ), avy Rev 812 187% (φανῇ--ὃ 95), 
ἐγχρῖσαι Rev 8:18 (ἔγχρισαι imper. mid.), ἐπιβλέψαι Lk 988 (ἐπίβλεψαι 
imper. mid.), τινά 1 Pet ὅδ in WH mg. (τίνα), Heb 512 (τίνα), ἢ ris Mt 19 
(i τις), ἀλλὰ Jn 628 (ἄλλα), ἐν Mk 48, 2° (WH ng. ἕν). 


Differentiation 
by accent. 


Syllabification. 


§ 31. The Greek rules for word-division, carefully 
laid down by grammarians, and normally observed in 
MSS and papyri, are important because of the light 
they throw on the syllabification of Greek speech. 
The most general rule is that the new line must begin with the largest 
consonant group which is capable of beginning a word. In addition to 
yp, Sp, 9p, KA, KB, Kp, KT, pv, TA, mp, mr, Tp, PO, xO, which are found 
initially, we have yp, Ou, rv and yp. A liquid (except in the case of pv) 
is taken to end the preceding syllable. On the same principle o is often 
thus abstracted, but here there were differences of usage: the verbal 
suffixes with o@ tended to bring the o over to the next line. Before a 
vowel o stood at the head of the new syllable. Proclitics were counted 
as belonging to the next word, and divided accordingly, except in the 
case of the prepositions εἰς, πρός, σύν and ὑπέρ. Similarly, if there was 
elision the two words ran into one. Thus we have xalr’ ἔτος P Oxy i. 101 
ter (A.D. 142), &]|s ἄν ἐδ. 11. 2708? (A.D. 94), παρ᾽ αὐτοῦ 10. °°, ἐϊκ rod P Lond 
2218 (=i. p. 7—B.C. 164), οὐϊκ ἐφρόντισας P Petr ii. 23 (3}}} (1i1/B.0.), μη δ᾽ 
ἄλλον P Reinach 1113 (B.c. 111), e/fo0=ex σοῦ Mt 2° &, ete. 

The rationale of this may be seen in English. If we used the same 
general principle of word-division, we should print “not a-t all,” for the 
syllables are divided exactly as in “not a tall.man.” For further illus- 
tration see Mayser 44 ff., also WH Jnér.? 315. 

There is a tendency to make continuous sounds, especially o and the 


Word-division 
and Syllabifica- 
tion. 


§ 31-32] SOUNDS AND WRITING. 61 


nasals, divide themselves between both syllables. The effect is some- 
times to double the letter, as év|yavrios, θαλίλάσσῃ (Mt 87° 8), προστάγ! 
ypacw, κατέσίσπαρκεν, pyno|orevdions (Mt 118 δὲ), sometimes to omit one 
of a genuine double, as ralowrnpias, ἄφεϊσαυτόν, ἐὰϊνῦν. The practice 
should be borne in mind as one calculated to generate various readings. 
Thus it might be applied to the reading πλήρης σῖτον Mk 478 Οὗ οι, 
which seems to be the oldest accessible: the alternative (see § 65) is to 
regard the adj. as indeclinable. 


Elision. 

§ 32. In the Hellenistic period the practice of 
writing elided final vowels (as in Latin) considerably 
advanced. In 1 Co 1538 the iambic line from Menander is written 
φθείρουσιν ἤθη χρηστὰ ὁμιλίαι κακαί, and there is no authority at all for 
χρήσθ᾽. This isa common practice in metrical inscriptions and papyri 
of our period. In prose the rules observable in the best NT uncials are 
set forth by WH (App.? 153) thus :— 

Elision takes place habitually and without 
variation before pronouns and particles; also 

before nouns in combinations of frequent occurrence, as am’ ἀρχῆς, κατ᾽ 

οἶκον. In other cases there is much diversity, and occasional variation. 

In ἀλλά elision takes place usually before articles, pronouns and 
particles, but with many exceptions and much variation. [In W the 
full form is normal (Sanders 25).] The passage Rom 614-83? js 
remarkable as having consecutively (with a single exception 7:5 
ἀλλ᾽ 6) 9 non-elisions attested by 3 or more primary MSS: in the 
six following cases (to 1016) there is no evidence for any non-elision. 
Elision is commonest before words (of all kinds) beginning with ε, 
rarest before those that begin with a. 

Δέ is never elided except in és δ᾽ ἄν, once or perhaps twice in 
τὸ δ᾽ αὐτό (not Phil 218), and perhaps in ἡνίκα δ᾽ ἄν 2 Co 316 (see 
margin) ; οὐδ᾽ occurs a few times. 

The places where WH regard the reading as open to some doubt may 
be tabulated thus: the figures represent the totals for the text, non- 
elisions standing first—the same figures reversed would accordingly 
represent their margin : 


Elision 


in NT uncials, 


ἀπὸ a. 0:2 dad. 0:3 ἐπὶ ἐ. 1:2 ἀλλά 28 : 38 
ὑπὸ a. 2:0 κατὰ a. 1:0 ἐπὶ ὁ. 1:0 δέ 1: 0 
ὑπὸ ἁ. 1:1 κατὰ ἑ. 1:0 ἐπὶ ὑ. 1:0 οὐξ 1: 1 
pera ἐ. 0:1 emi i. O:1 
pera ὁ. 1:1 


Thus, in places where MS evidence is evenly balanced, WH 
prefer non-elision 39 times and elision 50 times. 
: : A comparison may be instituted with some repre- 
In papyTl. sentative papyri. For this purpose nos. 1-41 in 
Milligan’s Selections are examined, dating from B.c. 311 to the end of 
11/A.D. 


62 A GRAMMAR OF NEW TESTAMENT GREEK.  [S 32 


(1) Prepositions.—Here elision preponderates greatly. “Emi occurs 
once unelided, and ἀπό three times, and all of these are with 
nouns. On the other hand we find elision with ἀπό (2-1 noun), 
διά (5-1 δι’ ὀλίων, 1 SV ἐγγύου), ἐπί (7-1 ἐφ᾽ ὕβρει, 1 ἐπ᾽ ἄχυρον 
1 ἐπ᾽ ἀληθείας), κατά (10-3 καθ᾽ ἡμέραν, 1 κατ᾽ ὄνομα, 1 κατ᾽ ἀρχάς, 
1 κατ᾽ οἰκίαν), μετά (3), παρά (8-1 ἃ name), ὑπό (5). The exx. 
not stated above are all with pronouns, and so suit WH’s rule. 
Nearly all the nouns are also in combinations answering to the 
requirement of “frequent occurrence.” These statistics may be 
supported by the totals in Witkowski’s collection of Ptolemaic 
private correspondence.! Here elision takes place 16 times where 
prepositions stand with pronouns, and twice with nouns (καθ᾽ 
ἡμέραν, παρ᾽ Ἰουδαίου). There is not one example of non-elision. 

(2) ᾿Αλλά has elision 3 times, and never non-elision; δέ 7 and 13 
respectively ; οὐδέ and μηδέ 4 and 4; re and ὥστε are never 
elided, nor are pe and oe. “Iva has elision in the formula ἵν᾽ 
ὑγιαίνῃς (so in Witkowski 8 times), but never elsewhere. In 
Witkowski ἀλλά is always elided (6 times), and δέ 14 times 
to 10; μηδέ 1:1, ὥστε 2:0, re 2:0 (unless τοῖς τ [ἄλλοις] in no. 
24 must count). Me and σε are not elided, except ἅ σ᾽ οὐ in no. 
46 (illiterate). 

(3) Rarely other words suffer elision : viz. ἠναγκάσμεθ᾽ ὑ., ἐποεῖσθ᾽ é., 
ἐνοφιλόμεν᾽ ἀ., μηδέν᾽ a. : add ὁπηνίκ᾽ ἄν from Witkowski. Ὅσας 
δ᾽ ἐάν might equally be read ὅσας δὲ ἄν, and is therefore not 
counted. The occurrence of μηδέν᾽ ἀπολελύσθαι in BGU i. 2714 
(= Milligan p. 101—1ii/a.D.) might be cited in support of μηδέν᾽ 
ἀπελπίζοντες in Lk 655, thus making ABL etc. agree in sense 
with the μηδένα of NW (cf. syr Yt Pesh), 


The bearing of these facts on the enunciation of 
sentences in the Common Greek is not unimportant. 
Elision is due to allegro utterance ; and the uniform 
absence of elision shows us under what conditions this was avoided. Aé 
and the enclitics re, με, oe, attached naturally to the preceding word, 
were not linked with the following word enough to produce elision. The 
prepositions and ἀλλά, being proclitic, suffered elision as naturally as the 
former did when compounded with verbs (with initial vowel) : excep- 
tions seem to occur almost only where nouns which are not parts of a 
formula may prompt lento pronunciation to produce a kind of emphasis. 

The subject is closely connected with one which 
figures largely in discussions of literary Greek, viz. 
the avoidance of hiatus which became an instinct with prose writers. 
Definitions may be found in Blass Grammar, 296 f., together with a 
detailed attempt to show that a partial observance of the hiatus rule can 
be seen in Heb. The attempt does not seem very successful, and no 
other NT book is accused of attention to this refinement. Dr. Rendel 


Allegro utter- 
ance. 


Hiatus. 


1 Documents also included in Milligan are omitted. 


ὃ 32] SOUNDS AND WRITING. 63 


Harris calls my attention to the avoidance of hiatus as a motive present 
in literary revisers of the text of NT, who would change the order of 
words for the purpose. This applies especially to the class of variants 
which Hort called “ Alexandrian.” But in genuine vernacular this dis- 
position to avoid hiatus was almost wholly absent. The well-marked 
tendency by which in vernacular Hellenistic hiatus is permitted in com- 
pounds at the juncture—as in rerpadpyns, ἀλλοτριοεπίσκοπος 1 ete.—may 
be set by the practice of writing elided vowels in verse, to create a pre- 
sumption that the later language was indifferent to the confluence of 
vowels. In pronunciation no doubt the usage was to sound the vowels 
rapidly, except in the cases where elision was still the rule, which means 
as we have seen cases of proclisis. Thus τετραάρχης was a real quadri- 
syllable, produced by the influence of other compounds in rerpa-, and 
χρηστὰ ὁμιλίαι (1 Co 15%) had the a o distinct, but pronounced in quick time 
so as not to disturb the rhythm of the verse. In this respect the difference 
between Hellenistic and classical usage is very much like that between 
modern and eighteenth century or older English: where poets used to 
write ‘th’ action,” we write and pronounce “the action,” though the 
scansion is the same. English dialects of course use the allegro forms 
very largely (as Irlams οὐ th’ Height, in thoos=in the house); and the 
analogy may prepare us for the probability that Hellenistic was not 
uniform. MGr implies as much by its free use of elision. 


Crasis. 


As we should expect, this result of virtual pro- 
clisis is greatly restricted in later Greek. In NT, 
except for τοὔνομα in Mt 2787, τοὐναντίον ter, and ταὐτά in Lk, crasis is 
confined to combinations with καί, which retains the same tendency 
in MGr more conspicuously. We find κἀγώ, κἀμέ ete., with which 
cf. κἀγώ, P Petr iii, 53 π.5 (iii./B.c.)—but καὶ ᾽γώ in P Oxy ii. 29418 
(A.D. 22); also κἀκεῖ, κἀκεῖθεν, κἀκεῖνος, and the stereotyped κἄν 
(MGr), on which see § 29 (7). Papyri of culture low enough to admit 
phonetic spelling show us that crasis was practised sometimes when 
unaccented words were capable of being fused with the preceding word : 
thus P Oxy iv. 744: (B.c. 1) ἐν ᾿Αλεξανδρέάσμεν, as we might write— 
σμεν with prodelision would suit our practice better, and means the 
same thing—, BGU iii. 975!" (4.D. 45) οὐλὴ καστρογνημίῳ ᾿κξ ἀριστερό. 
Note κἀπεθύμει in D* (Lk 1516). 


Crasis. 


Combinations written as one word. 


There is no evidence earlier than the minuscules 
One word or : ξ 
thie by which we may decide whether to print phrases 
5 like δέκα ἔξ, am’ ἄρτι, ἐφ᾽ ἅπαξ, ὑπὲρ ἐκ περισσοῦ 
etc. as one word or as two or three. The only case in which it matters 


1B have ἀλλοτριεπ., which may well be right, though papyri have parallel 
forms to support the other: see Vocabulary, s.v. 


64 A GRAMMAR OF NEW TESTAMENT GREEK. [§ 32-33 


—for of course in the older uncials there is no division of words—is 
where the fusion involves assuming that the accent of one element was 
dropped. Sometimes MGr may give possible evidence, though we must 
not lay too much stress on it: fusion may easily be of late origin, and 
the evidence of MSS which divide or punctuate off the words may be 
of equal weight, since though in danger of being literary it is of higher 
antiquity. Thus MGr ἀνάμεσα disagrees with ἀνὰ μέσον which Ti prints, 
presumably on evidence of MSS. 


Vowels. 


§ 33. The vowel system in the Κοινή has 
undergone more extensive change than is ap- 
parent from the spelling, which is still largely dominated by 
the literary tradition. The operative factor in change was 
the development of a stress accent out of a pitch accent 
(see § 30), which necessarily worked towards an ultimate 
levelling of quantity in vowels. In MGr long vowels and 
short are not distinguished: an accented syllable is half-long 
with an o, an unaccented one short with an w. The process 
which thus radically altered the whole sound of Greek 
speech was not complete for generations after the NT, but it 
had set in strongly, and must have done much of its work. 
In so far as accent was gathering stress character, it neces- 
sarily produced the ablaut effects which we may note in 
English, where also spelling fails to express many of the 
products of accentual conditions. Among those which find 
expression in writing we may instance “ Kretschmer’s Law” 
(see § 30) by which an unaccented short vowel after a liquid 
or nasal fell out when the neighbouring syllable had the same 
(or nearly the same) vowel. Thus Βερενίκη became Bepvixn, 
σκόροδον σκόρδον, ἐπηκολούθηκα loses its first o and 
ἀπελήλυθε its v. Illiterate papyri and inscriptions show 
more of these syncopations than “correctly” written docu- 
ments,! but doubtless ordinary speech showed them plentifully. 
They are like our own pronunciation of words like léboratory, 
where the first o is practically crushed out by the stress 
before it. The result is that Vowel-gradation (Ablaut), which 
ceased to work during the earlier period of Greek, when all 


Vowel System. 


1 But our uncials are not without instances: see a list in Thackeray 99 f. of 
such forms found in LXX. 


§ 33] SOUNDS AND WRITING. 65 


syllables had equal weight and there was no force operating 
to produce disintegration, started afresh as it must whenever 
stress accent comes in. It will not be necessary to enlarge 
upon the later gradation developments, for they only pro- 
duce visible results on a large scale when we come to the 
MGr: of course gradation in its prehistoric working belongs 
to the description of the earlier language, from which the Kowvy 
inherited the familiar series λέγω : λόγος, λείπω : λέλοιπα: 
λιπεῖν, πένθος : πέπονθα : παθεῖν, ἵστημι : ἴστᾶμεν, etc. Our 
only concern with <Ablaut here is to observe that it is 
not simply a force acting in the period when the Indo- 
European languages were undivided, but a necessary and 
constant sequence of stress accent, only suspended when 
language takes the musical accent like French or ancient 
Greek. 

It need only be remarked further that in 
i/A.D. itacism was levelling the vowels con- 
siderably. Ac and ε were not far behind in their fusion, and 
οἱ and uv followed in time: see each development discussed 
below. These changes of pronunciation are of great import- 
ance in textual criticism. The extent of itacism in an early 
uncial may be well illustrated by a summary in Sanders’s 
introduction to Codex W (p. 20):— 

In the first quire of John the itacisms are as follow: . for εἰ, 193 


times ; εἰ for s, 17; ¢ for a, 82; a for ε, 16; ο foro, 3; .fore,3; o for 
w, or ot for ἢ, v for οι, « for n, and ε for v occur once each. 


Itacism. 


Of course some of these are mere isolated freaks: the nature 
and significance of the rest will appear from the following 
pages. 

We proceed to take the vowels seriatim, starting in 
each case from the classical form and chronicling variations. 


iA 
(a) Short. 


There are some cases of substitution of ε for & in 
contact with p. The possibility of phonetic origin 
must be considered, especially as it is accepted by Thackeray for some 
LXX phenomena of the kind. The Egyptian deity Serapis came into 
Greek first as Σαρᾶπις (Mayser 56 f.); but Mayser gives two Ptolemaic 
instances of Σεραπιεῖον, and suggests that the distance from the accent 
was responsible, coupled with the influence of p, S<pams does not appear 

VOL, Il—5 


A short. 


66 A GRAMMAR OF NEW TESTAMENT GREEK. [8 33 


till the Roman age. ’Epoevixds also (Mayser 5) seems best explained by 
the two forces which produced Σεραπιεῖον : an isolated περά for παρά 
(P Tebt i. 110*—8.c. 92 or 59) may be thrown in. Τέσσερα is witnessed 
first by a British Museum inser. from Egypt (p.c. 51-47), OGIS 1931! 
radav[ta téo]oepa ;! next comes BGU i. 133° (a.p. 144)—there seem to 
be hardly any others till the Byzantine age.? Téooapes as accus. 
belongs to an entirely different category : see the accidence, 8 71 (a) (y). 
The case for τεσσεράκοντα is distinctly better, but it is greatly 
outnumbered by the a form until the Byzantine period.* It is perhaps 
significant that the earliest instance we possess (see below) is of the 
ordinal, where the ap syllable was further away from the accent than in 
the cardinal. When we note that four differed from forty in the fact that 
the accent preceded the ap syllable instead of following it, we have a 
reason for presuming phonetic causes at work. Ionic influence * would 
not account for the unequal treatment of parts of the same numeral. It 
is significant that τεσσάρων maintained itself even in Byzantine, as in 
MGr; ef. P Flor i. 87 (v/vi Α.}.), where τεσσάρων and τέσ[σ]ερα come in 
successive lines. That the order of development is seen in the chrono- 
logical succession Σεραπιεῖον and ἐρσενικός and τεσσερακοστός, then 
Sepamis and τεσσεράκοντα, and finally réooepa(s) much later, encourages 
us to hold that the earliest change depended on the position of ap well 
before the accent. The evidence here given will suffice to make it 


1 Or δέκα [τέσ]σερα, as given by Strack, Archiv i. 209. 

2See CR xv. 33a, also xviii. 107a, where I cite εἰκοσιτέσσερας bis from a 
ii/A.D. papyrus : CPR 242 (cited in the former paper) is apparently a mistake. 
Mr. Thackeray (in a letter) retracts his statement (p. 74) that ‘‘ Σερᾶπις and 
τέσσερα appear to have come into general use together about i/a.p,” Add 
P Oxy viii. 11424 τέσσερα (late iii/a.D.). 

3 For τεσσεράκοντα before ii/A.D. may be cited P Tebt 11. 3888 10 (a.p. 98), 
P Gen 24" (a.p. 96), CPR 220! (i/a.p.), P Flor i. 616? (a.p. 86-8), ib. 86° 
(i/a.D.), BGU iii. 916* (Vespasian), P Lond 262) τεσσερακοστοῦ (A.D. 11=ii. 
p. 177): the same document has τεσσαράκοντα bis. Mr. Thackeray’s tables for 
papyri published before 1907, which he has kindly sent me, contain 25 
instances of τεσσεράκοντα (-κοστός) from i/ to iii/A.D., and 46+with a. In 
making my own tables I have ceased to enumerate instances of a for the early 
centuries. That there was a marked difference between cardinal and ordinal 
may be seen well in P Flor i. 1 (A.D. 153), where τέσσαρες [nom. and ace. ] 
occurs twice and τεσσεράκοντα four times. See also Prol.? 243f., and Mayser’s 
reff., p. 57. Ihave one ex. of τέσσερᾳς, a Tebtunis ostracon of 8.0. 6/5 %, in 
P Tebt ii. p. 337. 

4 Which accounts for τεσσεράκοντα in the 1i/A.D. Homer, P Tebt ii. 265 
(11. ii. 545), and the Teos inser. Sy/d. 177” (B.c. 303), although the latter is in 
Kow7 Greek. 

51 can quote only one instance of accented a>e,viz. BGU iv. 1013° (mid 
i/A.D.) θυγατρέσι, which is not a case of ap but of pa: moreover the case may 
be one of blundering declension, the -τερ- of the stem affecting the abnormal 
-rpa-. Mayser’s instances (p. 58) are referred by him to Coptic influence. 
KpéBarros (twice in W), which has left its mark on MGr, is too late to be 


brought into the case. 


§ 33] SOUNDS AND WRITING. 67 


certain that τέσσαρες etc. and τεσσαράκοντα (-κοστύς) were the autograph 
forms for NT. Against these stand “Jn 195 τέσσερα SNALM, Rev 46 A, 
45 SA etc.” (Debrunner), with τεσσεράκοντα regularly in the uncials : 
see Ti.-Gregory 80. Dissimilation, which WH (App.? 157) postulate as 
cause, is excluded by the fact that τέσσερας has no attestation except 
A once in Rev 4. On τέσσερα it is only necessary to observe that 
sporadic instances meet us in documents contemporary with Jn and 
Rev. In proper names of foreign origin there is considerable variation, 
but it need not occupy us where Semitic is concerned. Δελματία 2 Ti 41° 
A (ep) Ο 424** cul? is compared by Deissmann (BS 182) with δελματική 
in BGU i. 937 (ii/iii A.D.), against CPR 2116 (a.p. 280) : Radermacher 35 
assigns it to Latin influence, which is likely enough. See Vocab. s.v. 
Δαλματία. Ildrepa, however, found in AC at Ac 211, has no European 
support. A different matter is the change of ap to ep in augmented and 
reduplicated forms of καθαρίζω. In LXX (Thackeray 74) A has it 14/21 
times, B once, while δὲ never shows it in LXX or NT. As we might 
expect, the evidence is scantier (and probably insufficient) for Luke and 
Heb, more satisfactory for Mt and Mk. Thus Mt 83 has ep in 
B*YELX*, Mk 14? in AB*CGLAII*al.; but Lk 45 ACLX, 17!4 AX,17 ALA, 
Ac 1018 ACLP 33 81, 119 AHL al., Heb 10? AC. Thackeray attempts a 
phonetic account of the change, but the fact that in one case (ἐκαθέρισα) 
the vowel is accented makes it much more probable that the ε is a double 
augment ; Schmiedel (WS 50) seems inclined to this. Of course it is 
abnormal, as a verb καθ-αρίζω, if a compound of κατά, would have made 
τήρισα. But this does not seem a serious objection. 

There are one or two other instances of ε substituted for an older a. 
᾿Εγγαρεύω, derived from the early (Persian) loan word éyyapos, appears 
in Mt δ: δὲ, Mk 15?! 8*B*: the spelling is not infrequent in inscrr. and 
papyri—see Vocabulary, s.v. It did not however ultimately prevail, as 
MGr ἀγγαρεμένος shows ; but its claim to a place in the autographs is 
strong, especially in Mk. Χλιερός stands in Rev 316 δὰ, ὕελος in 
Rev 2118 cu?! (two have AA) and so 2121 46 15? bis. For LXX forms 
of this kind see Thackeray 75: the variation was said to be dialectic, 
ὕαλος and χλιαρός being Attic. See Lob. Phryn. 282 (Rutherford NP 
364), Moeris 418, Thumb Hellen. 75f., Schweizer Pery. 361. (where 
inscriptional evidence is cited). Thumb ({.6.) thinks that both’a and ε 
forms had their place in Hellenistic. In NT χλιαρύς, ὕαλος (ὑάλινος) and 
φιάλη seem certain, though WH admit χλιερός as alternative (App.? 157). 

The variation between ἕνεκα and ἕνεκεν, εἵνεκεν is due to dialect 
mixture. The Attic ἕνεκα in Ac 26”! is regarded by Blass as in keeping 
with a speech in the presence of royalty: it appears also in Ac 1953 
NAB, Lk 623 (exc. DW al. pauc.), Mt 195 (exc. CDW), Mk 139 B. Ἕνεκεν 
(partially Ionic in origin) is normal, but the still more Ionic εἵνεκεν 
figures in Lk 4:8 1829 SB, Ac. 289° N*¥A, 2 Co 3! (exe. C al.). Ἕνεκεν is 
the normal form in later Attic and Κοινή : see Mayser 241 f., Crénert, 


1Mr. H. Scott points out that in Test XII Patr. Charles (Oxford, 1908) 
gives following v.d/. Reuben iv. 8 ἐκαθάρισεν (his text) af, ἐκαθάρησεν c, 
ἑκαθαίρισεν Ὁ, ἐκαθέρισεν de. 


68 A GRAMMAR OF NEW TESTAMENT GREEK. [ὃ 33 


Mem. 113f.—the two other forms occur relatively much as in NT. 
The similar Ionic εἶτεν is only found in Mk 478 xB*LA: Phrynichus 
gives it and ἔπειτεν a specially bad mark (ἐσχάτως βάρβαρα), and a 
solitary appearance in the least cultured book of the NT is quite in 
keeping. 

A has yielded to o in 1 Ti 1° πατρολώαις, μητρολώαις, Where only cursives 
have pa (exc. K πατραλ.). The consciousness of origin (marp-adoias, cf. 
ἀλοιάω) has weakened, and analogy of πάτρο- and μητρο- compounds 
prevails. Barradoynonre Mt 67 NB (W βατταλογεῖται) is replaced by 
βαττολ. in the other MSS (βλαττολ. in D*). See Vocab. s.v. and below 
§ 105. Since the word is probably for Barrado-Aoy., by haplology, the a 
is original, and Barrod. will be due to wrong association with Barros. 
Mayser 60-62 gives a few exx. of a>o and o>a, but they are not enough 
to be significant. 


(b) Long. 


Sivame shows the Ionic ἡ in Lk 176 W. This 
A long. 


would have been the Attic form if Attic had used 
the word : at Athens they said νᾶπυ (Lob. Phryn. 288, where Κοινή cita- 
tions for σίνηπι are given). The survival of a in λαός and ναός is noted 
in ὃ 34. Mayser thinks λαός may be originally a poetic word, used 
primarily in plur.: see p. 29, but note also Thumb’s comments on this 
section in Archiv iv. 490. The flexion of the -j7@ verbs, with their 
tendency towards the -aw type, is discussed under verbal accidence. In 
two words apparently the Ionic 7 has secured a place in the Κοινή, 
διηνεκής (Att. diav., but not consistently), and πρηνής (usually taken as 
=Att. mpavns): if Chase’s explanation of the latter is right, the root is 
pré ‘burn.’? So also χορηγεῖν (Att. yopay.), which is not so much in 
Tonism as an effect of analogy (στρατηγός, ὁδηγός). It is needless to bring 
in proper names, which might of course start in Doric or other dialects 
as easily as in normal Κοινή. 


2. E. 


Two verbs in -a¢w appear in NT where e would be 
expected. ᾿Αμφιάζει replaces the literary ἀμφιέννυσιν 
(still in Mt 680) in Lk 1278 B, where the rest read ἀμφιέζει : so ἠμφιασμένον 
Mt 118 D, and predominantly in LXX (Thackeray 75). The outside 
evidence for -d¢w is not early: see Vocab. s.v., and add P Iand 6214 
(vi/A.D.). Radermacher 35 accepts the explanation of it as coming 
directly from ἀμφί, as ἀντιάζω from ἀντί. ‘The fact that ἀμφί did not 
survive in the Κοινή is one of the difficulties in this view. I prefer the 
account which Schweizer, Perg. 37, takes from W. Schmid, that -d¢w (said 
to be Doric) is the product of a very large class of -d¢# verbs, which 
naturally exerted strong influence on the comparatively few verbs in 
-ἐζω. The new present ἀμφιέζω was of course a back-formation from the 


E and a. 


1 See J7S xiii. (1912) 278 ff., and J. R. Harris in AJT xviii. 128 f. 


§ 33] SOUNDS AND WRITING. 69 


aor. in -eoa. Πιάζω may be similar in history, but it seems to have 
assumed a new meaning, which would encourage the supposition that. it 
came into the Κοινή separately from some dialect : Theocritus 43° πιάξας 
τᾶς ὁπλᾶς, “gripping it by the hoof,” is warrant for the Doric provenance. 
The older form still survives with the meaning “press” in Lk 688, 
MGr πιάνω “take” joins several papyrus exx. in support of the Κοινή 
mato with that meaning. 

The interesting blunder ἔλαβον γυναῖκας in Heb 11% p!3N8*AD* is 
perhaps not phonetic in origin, though Thackeray 149 quotes three exx. 
of nom. in -as from LXX ; and the opposite, γυναῖκες for -as, occurs in 
P Catt (Chrest. ii. 372" *)—cf. § 55(4) on ace. pl. in -es. The import- 
ance of the reading consists in the link it helps to establish between 
all the best MSS: the mistake was very probably in the autograph 
itself. 

The itacistic variation between ε and ai figures 
incessantly in the MSS, as in papyri and other wit- 
nesses. It was indeed hardly even a case of “shortening” (as WH 
App.” 157f., where στύλος and κρίμα are put on the same footing): 
e and a were as completely identical as « and εἰ in the uncial 
- period. WH 309f. give Gal 418 as “one of the few instances in 
which B and & have happened to fall into the same itacistic error,” 
(ηλοῦσθε for -a. Practically this means that where either would make 
sense we are allowed to choose for ourselves on other considerations 
than MS authority. When in 1 Tim 6, 2 Tim 216 some 6-text 
authorities (FG and a few cursives) read καινοφωνίας for κεν., we 
recognise it as a legitimate interpretation of what when read aloud is an 
ambiguous word. But our acceptance of it is discounted by finding that 
the version which adopts it (lat vt (vs seme)) js one which has a close nexus 
with the 6-text uncials; while the Sahidic, Bohairic and Syriac vss. 
justify the spelling with ε. But in Mt 1116 the choice between ἑτέροις 
and ἑταίροις is not so easy, in spite of WH’s description of the latter as a 
“perverse confusion” (p. 310). Ἑταίροις is read by poor uncials only, 
but the Old Syriac (*" and 1) and the Latin Vulgate support it ; and that 
in many uncials (including CL) ἑτέροις means ἑταίροις, is shown by the 
addition of αὐτῶν. Indeed δὲ itself makes us uncertain as to its testimony, 
when just above τοῖς ἑτέροις we read τὲς ἀγορᾶς, with a in each case 
supplied by the corrector ; can we say more than that the corrector saw 
nothing wrong with ἑτέροις, while he did with dyopés? The oldest 
tradition in vss. is here divided, the Old Latin being against the Old 
Syriac. The rendering of WH (“the other ‘side’ or party in the 
game”) is very attractive ; but I should accept it on its merits, and not 
because in such a matter NB are to be preferred to CL. 

In one frequent category a and ¢ involve different renderings, that 
of inf. -σθαι against imper. -o6e, as in Lk 14!7,19!% There is further 
the choice between ἐπάναγκες (most MSS) and ἐπ᾽ ἀνάγκαις NAC in 
Ac 1578 ; the second does not seem to have much of acase. Only one 
remark need be made as to words which through isolation may have 
lost their traditional spelling, such as κερέα, ἐξέφνης etc. (WH App.? 158). 


E and au. 


70 A GRAMMAR OF NEW TESTAMENT GREEK. [ὃ 33 


The acceptance of ¢ here is justified if it is clearly understood that the 
preponderant spelling of the oldest uncials is taken simply as the best 
attested, and is not claimed for the autographs. For these the papyri of 
i/ or 11/A.D. are decidedly better evidence. Thus for φελόνης 2 Tim 41% 
SACDEFG I know only one papyrus parallel (P Fay 347, 1ii/a.p.) 
against twelve occurrences of dav. (par. ter) from papyri of i/-iii/.! 
Similarly αἰφνίδιος has one papyrus parallel (ev. none—see Vocab. s.v.), 
while ἐξαίφνης is less clear.2 ’Emavayxes (which Blass called doctum voca- 
bulwm !) is very common in papyri with this spelling, and there is nothing 
to suggest the alternative. So far then as the evidence of our vernacular 
documents goes, there is little encouragement for deserting the traditional 
orthography in words where the substitution of ε does not affect the sense : 
the variation appears to be purely casual, and probably nowhere became 
a habit as early as i/AD.* On the date of the development see below, 
§ 36, under αι. 

Schmiedel gives the following list of ambiguous places where we must 
decide between ε and a on internal grounds. ἔγειρε Mt 9° al., ἀνάπεσε 
Lk 14!° 177, παρένεγκε Mk 1438, Lk 224, ἔρχεσθε Lk 141’, πραγματεύσασθε 
Lk 1918, φυσιοῦσθαι 1 Co 4°, ἀγνοεῖτε 1 Co 14°8; also the choices in 
Mt 1116 and Ac 1558 mentioned above. (See WS 471.) 

It will be shown later (§ 36) that the contacts 
between ε and εἰ are of a special nature, and do not 
fall into the same category as those with « proper. One conspicuous 
instance of the latter is the series of compounds in dpy.-, which becomes 
a very large one in Hellenistic. Forms like *“ApyéAaos—to name 
one which survives as a proper name in NT —are older; but 
the change is not phonetic, but extended from the influence of a 
special category found in early Greek. Instances of « for ε in Egypt 
may sometimes be due to the native language, where e and 7 were 
very close together. See Mayser 80ff. That ἀποστείλω in Ac 7** is 
misspelt for ἀποστελῶ is wrongly assumed by WS 43 n.?7: it is a 
matter of syntax, not orthography. See Prol. 185, and additions in 
Einl. 292. 

A special case is dvayaov, which in Attic presumably would have 
ended in -yeov : see Rutherford NP 357f. The Hellenistic form was 
Doric or Tonic. 


E and ει. 


ΤΡ Oxy ive 198 ter (6. 4.D. 1); P Gigs 103, 12%, 79-3, P Oxy i, 5511: 
vi. 933°9, P Hamb 10! (ii/A.p.); Βα iii. 81674 and P Oxy vi. 936 bis 
(i11/A.D. ). ἊΣ 

2’ Εξαίφνης in P Par 518 (Β.σ. 160) and in P Flor ii. 1757 (A.D. 255) ἐξεῴφνης, 
where the correction is significant. Per contra ἐξέφνης in P Giss 867 (early 
ii/A.D.), and ἐξεφάνης in Cagnat 111. 1145 (Syria). 

3 How little significance the variation has in the uncials may be seen from a 
random page of δὲ (Mt 9*°-10!7), where out of 23 occurrences of a no less than 9 
are wrong, while 6 times ε is wrongly written for a. On the other hand, in 
Rom 3°—4" there is but one itacism (aivds) to 40 which are right. 

4 Add there P ΒΥ] ii. 2338 (ii/A.D.) dvevéyxw δὲ ᾿Ηρακλείωι, “1 will refer 
it to H.” 


§ 33-34] SOUNDS AND WRITING. vet 


The change of e to n»—for which see below, ὃ 34— 
may be tentatively recognised in the special case of 
πλήρης for πλῆρες, Which is not uncommon in papyri of the Roman age. 
Thackeray 176 f. thinks this may have arisen partly from the working of 
the p flanking two vowels that were approximating in sound. This 
would first assimilate πλῆρες and πλήρεις, and then πλήρη, with πλήρους 
following later, so that the adj. became indeclinable. The explanation 
is not without difficulties, but is perhaps the best yet offered. On the 
appearance of this indeclinable πλήρης in NT see under the flexion, § 65. 

E has been assimilated to ο in ὀλοθρευτοῦ 1 Co 1010 
(all but D*), ὀλοθρεύων Heb 1178 (Nw, against ADE), 
ἐξολοθρευθήσεται Ac 355 (NEPW, against AB*CD). Thackeray (p. 88) 
shows that it was rare in LXX uncials: it was a later development, due 
to the A, and surviving in MGr ξολοθρεύω. He compares the much earlier 
development of ὀβολός in Attic out of ὀβελός, on which see Meisterhans 3 
22 n. In ὀχύρωμα (2 Co 104, P Petr ii. ter—see Mayser 96) we have a 
variation from Attic, which had ἐχυρός and ἐχυρόω ; but ὀχυρός was older, 
and Attic seems to have assimilated to ἔχω. 


E and ». 


E and o. 


§ 34. Attic ἡ maintained itself in the Κοινή over- 
whelmingly, against the a of all dialects except Ionic. 
This, the most impressive evidence of the predominance of Attic as basis of 
the Κοινή, is confirmed by the reversion of ἡ when “ pure” to a, wherein 
Ionic differs from Attic (lon. πρήσσω, ἀληθηΐη etc.). (The Kown naturally 
never changes pre-Greek ἡ, as in τίθημι: Elean was the only dialect which 
did so.) Compound nouns in -ηγός (from ἄγω) not only show the 7 
throughout (ἀρχηγός, ὁδηγός, στρατηγός), but their analogy changes the 
Attic χορᾶγός, so that in the Kowwn we have only χορηγέω. Aaos and vads 
(contr. νεωκόρος) established themselves in Hellenistic from dialects out- 
side Ionic-Attic. There are sporadic occurrences of a forms. For ὁδηγ. 
we find ὁδαγοί and ὁδαγῇ Mt 1514 D, ὁδαγεῖν Lk 689 D, and ὁδαγήσει Ac 
851 B*: B shows the a 1/8, D 3/7. In Mk 918 D 565 read ῥάσσει, which 
need not be equated with pyooe—see below, ὃ 95: the a is therefore 
short. The very attractive reading of B in Ac 27?" προσαχεῖν (g resonare), 
which accounts for the variants, has the difficulty of being a Doric (etc.) 
form which disagrees with the common derivatives of the same root: 
κατηχεῖν and ἦχος are conspicuous in NT. Can it have been a term used 
by sailors from Crete, Cyprus, Lesbos, Corinth, or some other maritime 
country outside the Ionic-Attic area, appropriated as a ¢.t.? One other 
abnormal a is that in Νύμφαν Col 418 ; in Prol. 48 it is argued that this 
was probably Νύμφᾶν, a woman’s name, not Νυμφᾶν masc., which involves 
a less probable reading (αὐτῶν). 

In two words class. ἡ is replaced by e. The 
spelling εἰ μήν cannot be due to confusion with εἰ 
“if.” as is shown by the quotation from a Doric inscr. in Prol. 46 η.΄. 
Thackeray (88f.) supplements the account there by observing that 
papyrus citations for εἶ μήν begin with B.c. 112: he also shows how 


Η. 


Η and ει. 


72 A GRAMMAR OF NEW TESTAMENT GREEK. [8 34 


LXX usage is sometimes inconsistent with any connexion with εἰ, though 
there are many instances where δὲ ON is rendered by εἶ μήν, and confusion 
with εἰ μή would be possible were it not for the plentiful papyrus exx. ; 
cf. Mayser 78n. To these may be added now some parallels. In Archiv 
v. 232, Wilcken quotes a papyrus of Wessely’s, which he reads εἶ (=) ov 
[δί-Ἰδοταί pou κιτ.λ. (reign of Augustus). BGU iv. 11418 (B.c. 13) ἡ ἔστιν 
ἢ ov gives us ἢ for εἰ: P Hamb 4’ (Α.Ὁ. 87) has εἶ μήν followed in 18. by 
εἰ (=7) ἔνοχος εἴην τῷ ὅρκωι, and so P Lips 121!8 (a.p. 151). In P Oxy 
viii. 1148? (i/A.D.) Hunt takes εἶ βέλτιον as for ἢ, asking a question: may 
this be simply the interrogative εἰ common in NT, and could we regard 
this use as originating in ἢ ἢ See Vocab. s.v. εἶ μήν for further instances. 
We may note that in 2 Co 2° we have to make this same choice with result- 
ing difference of meaning—e (1.6. ei) with NCDGw, or ἡ (=7) with AB 33. 
We find one or two exx. of this itacism, earlier that our uncials, in two 
Oxyrhynchus MSS: P Oxy iv. 655!4 (uncanonical Gospel, not later than 
A.D. 250) εἱλικίαν, and 656% (LXX of Gen 1955, %—early ili/A.D.) εἴ]δη, 
εἴδη, for 75e. See further Mayser 78f. He puts under this heading the 
new perf. τέθεικα for τέθηκα : the latter stands alone in Attic inserr. from 
400 to 200 B.c. (Meisterhans* 189, Schweizer Perg. 184): τέθεικα begins 
to appear in ii/B.c., as Perg. 24875 (B.c. 1385/4). The old explanation of 
this from analogy—ddjxa : ἀφεῖκα : : ἔθηκα : réOecka—seems however to 
be adequate. But no analogy is apparent for κηρίαις Jn 114* AX al., nor 
for avarepos Lk 147! ABDE*LW (-πιρ. SPR) al., which is further 
recommended to us by the stigma of ἀμαθία in Phrynichus (ap. Ti in 
loc.). His note mentions further that the Hellenistic spelling was εἰ 
and not ε.} 

In the same category stands. for ἡ. So Rev 1812 
σιρικοῦ (all uncials), for which WH quote C7G 5834 ; 
1 Co 4" γυμνιτεύομεν (all exe. L— -νειτ. B*D*) ; σιμικίνθιον Ac 1917; 
Κυρήνιος Lk 2? (exc. BW)—Blass thinks that Κυρήνη provided the sugges- 
tion; ov pt [μνησθήσο]μαι Heb 10!” (013); πλήρις Jn 114 (W). The 
name Μιτυλήνην Ac 2015 ends with -ίνην in AEL. Under this heading 
comes also Lk 23!! W ἐξουθενίσας, which however may be a deliberate 
correction, since it agrees with the form that seems to have been used by 
Plutarch : see below, ὃ 46, and further on this verb ὃ 119. The spelling 
Χρηστιανός &* 3/3 (also 81 in Ac 1156) is not so much an itacism as a 
consequence of the common pagan misreading of the mysterious name 
Χριστός. Finally, we may preserve for the museum of exegetical curiosities 
the Byzantine invention of a κάμιλος “cable” to be an improvement on 
κάμηλος in the paradox of Mt 1935 and [15 (Suidas and a scholiast). 
Instances of « for ἡ in LXX may be noted in Thackeray 85: they are 
“distinctly rare in B and not much commoner in SA.” See above, ὃ 28, 
for Rudberg’s note on the rarity of a confusion of y and. in ἢ. 


H and tu. 


1 WH compare ἄπειρος (gs. ‘‘unmaimed”’) in Hdt. i. 32: if the reading is 
sound, we might take the NT word to be Ionic. But cf. Stein in loc. It 
should be added that in Heb 1151 WH mention with some approval the con- 
jecture ἐπηρώθησαν for ἐπειράσθησαν, which would come by way of ἐπειρώθησαν. 


§ 34] SOUNDS AND WRITING. 73 


The very common MS confusion of ἡμεῖς (and 
cases) with ὑμεῖς is discussed under v in ἃ 35, and a 
suggestion is made as to the phonetic cause. WH Jntr. 310 urge that 
confusion of sound plays only a part: they recognise a “ prevailing 
tendency . . . to introduce ἡμεῖς wrongly, doubtless owing to the natural 
substitution of a practical for a historical point of view.” 1 Pet and 
2 Co are specially mentioned as largely affected, and Ac 1738 [B cu® boh] 
τινὲς τῶν καθ᾽ ἡμᾶς ποιητῶν is noted as a reading they would much like 
to accept, but for “the limited range of attestation.” 

Nouns in τῆμα are sometimes found with -eya, 
through the strong association with the -ovs class, in 
which the weakened root-vowel was primitive: εὕρεσις, θέσις etc. pro- 
duced εὕρεμα, θέμα, just as κρίσις produced κρίμα. See § 30, and 
Thackeray’s list, Gr. 791. On ἐρρέθην, ῥηθῆναι, see ὃ 95 s.v. cipw. In 
verb-endings WS 48 gives στήκετε Mk 1175, 1 Thess 3%, θαυμάζετε Jn 5?°, 
ἐξομολογήσεται Phil 211, as instances of well-supported but grammatically 
dubious substitutions. 


H and u. 


H and ε. 


An apparent extrusion of ἡ after οἱ is rather widel 
Hidropyed. found sas Poin of ποιέω : cf. also βοιθοῦ for ols) Bob 
in Wilcken Ostr. 10841! (B.c. 136), 10897 (B.c. 135). So ποῖσαι Lk 1142 8, 
ποίσας Jn 511 W. Mayser 83 cites four papyrus exx. from ii/B.c. See 
Blass Pron. 38n.5, and G. Meyer Gr.? 132n.?; also Radermacher 34, 
who accepts Hatzidakis’ postulate of a parallel stem ποίω, like νόω 
against νοέω. 


4. Ο. 


Α replacing of o by a is seen in μεσανύκτιον (-ου) 
Mk 13% B*W, Lk 115 D*. ΜΟῚ μεσάνυχτα shows 
that it is not a mere accident: the adv. μέσα may suggest its origin. 
Aipapootoa Mt 9?° δ is an obvious assimilation to the primary. 
Κολασσαεῖς in the title of Col in AB*K al. and in δὲ twice at the top 
of the page, is shown by Lightfoot (Comm.’ 17n) to be a later spelling : 
Paul’s own spelling in 1? is Κολοσσαῖς. 

The close approximation of o and , which has 
plentiful evidence in ii/B.c. in Egypt, but in Attica 
not before ii/a.D. (Thackeray 89), accounts for some doubtful spellings, 
and for a few more important matters. WH mention the follow- 
ing :—Svxopopea (LE 19* DQ -pop.); χρεοφιλέτης (Li 1», U 3.» have the 
“correct” χρεωφ.) ; πρόιμος (from πρό, antithetic to ὄψιμος) but πρωινός 
(from πρωΐ), both as in LXX—see Thackeray 90; ἐνδώμησις, now con- 
firmed by the inser. Syll. 583%° (1/A.D.)—o is only due to false etymology ; 
Στωϊκός, in Ac 1718 B, is the original spelling (see Brugmann-Thumb 
64), which in SADE al. is changed to Sro. by influence of στοά. Nouns 
in -ωσύνη and comparatives in -wrepos may be taken as maintaining 
their place when the preceding syllable is short: such exceptions as 
there are explain themselves simply by levelling, which works easily 
when the two sounds are getting very near together. 

The fact that the uncials generally preserve the historic spelling in 


O and a. 


O and w. 


74 A GRAMMAR OF NEW TESTAMENT GREEK. [ς8 34 


cases like these where nothing turns upon it must count for something 
in their favour when it is urged that mere itacism accounts for the 
change of -ομεν into -wpev, or the like, which WH (309) call “ probably 
the commonest permutation.” It may be questioned whether the great 
uncials show instances of ὦ for 0, or o for w, in which the matter is one of 
mere orthography. Thackeray (91) remarks, “In the LXX at least we 
shall not expect ἔχομεν and ἔχωμεν to be confused in Cod. B”: the 
rarity of this confusion in the principal LXX uncials comes out strik- 
ingly on p. 89f. He is applying this to the crucial case in Rom 5}, 
The only objection to following the uncials here is removed by syntac- 
tical considerations, which will be taken in their place: meanwhile see 
Prol. 110, 247. Here we find all the uncials (exc. S@GP) and all the 
vss. on the side of ἔχωμεν ; so also patristic testimony from Marcion 
down, except Didymus, Epiphanius, Cyril (3). If then Paul really 
meant ἔχομεν, We must assume a primitive error in the written text, due 
possibly to a mishearing on the part of his scribe. Nachmanson Magn. 
64 ἢ, gives inscriptional instances of confusion of o and ὦ in Asia, especially 
Hierapolis. But it can hardly be called common in Asia Minor in i/A.D. ; 
and the fact that in Egypt the fusion was practically complete before 
Paul’s day proves nothing for our purpose, since Egypt was manifestly 
ahead of other countries in this change. Schweizer Perg. 95 gives further 
evidence for the Asia Province, but it is also indecisive. Champions of 
ἔχομεν must first examine the point of syntax, and then if dissatisfied 
claim the right to emend the text: their case cannot rest on authority, 
nor on the assumption of itacism.! A further important instance of the 
same confusion is 1 Co 1649 φορέσωμεν, if we are to listen to the modern 
authorities who disapprove of the hortatory sense which Tertullian and 
Chrysostom attest. Here B and one cursive support o: it is probably 
again a matter of syntax and not itacism, Jas 415 ζήσωμεν is best taken 
as a genuine v./. due to misunderstanding of the construction. To take 
exx. of another kind, we are justified in regarding καθαρίζων in Mk 
719 NABLAW al., μεῖζον in Jn 1055 AB*X (μείζων in MSS reading ὃ 
being a compromise), and many other readings where the question is 
between o and o, as genuine vv.ll., and not mere misspellings. 
Perhaps καυθήσωμαι 1 Co 195 CK al. κερδηθήσωνται 1 Pet 31 cu, stand 
about alone as genuine exx. of the itacism which we are asked to recog- 
nise in Rom 5! and 1 Co 154%. Schmiedel (WS 48) gives sundry other 
verb forms where o is well attested for what he regards as a grammati- 
cally necessary ὦ : viz. Rom 1419 διώκομεν, Mt 1915 al. ἰάσομαι (LXX), 
1 Jn 52° γινώσκομεν, Ac 2154 Evpnoovra, Rev 1415 ἀναπαήσονται (!), 
Gal 613 διώκονται, Rom 148 ἀποθνήσκομεν, Lk 3!4 ποιήσομεν, Jn 41% 


11 modify, mainly on Thackeray’s lead, the freedom I claimed for the 
modern editor in Prod. 36. I haye H. A. A. Kennedy (cited Prol. 247), 
Rendel Harris (Zxpositor, vitt. viii. 527) and Deissmann (Paul 148) against me 
—a trio I am not likely to ignore. Nor do I object to their use of virtual 
conjecture ; I only cannot see why it is necessary, when the sense is practically 
the same. 


§ 34-35] SOUNDS AND WRITING. 75 


διέρχομαι, Rev 611] ἀναπαύσονται; also vice versa Lk 918 ἀγοράσωμεν, 
Mk 657 δώσωμεν. He mentions further two tempting conjectures, 
Rom 13? ἀγαθοεργῷ and 2 Pet 3° δι᾿ ὅν. In 1 Co 957 Lk 185 ὑποπιάζειν 
is an unsuccessful attempt to explain an unusual word, and not an ortho- 
graphic variant. 

Corroboration of the general line here taken may be found in the 
fact that only two of the iv/a.p. and earlier Biblical texts in P Oxy i.-xi. 
(see 8 36) show this itacism, viz. ii. 209, where in Rom 15 ὅ we find 
πρωφητῶν and ὑπακωὸν πίστεος (GH call this document (iv/A.D.) a 
“schoolboy’s exercise”) and iv. 657147 (=Heb 10%, iv/a.D.) κρίσσωνα. 
In W I can see no instance, unless we count παιδίον for -ων in Mk 957, 
and even this is translatable and may be a genuine v./. Scrivener (Colla- 
tion of Codex Sinaiticus (1864), p. 1111) says that δὲ interchanges ov and ὦ 
20 times in the termination of verbs, and o and @ 68 times. It soon 
appears that the question is one of syntax as much as of orthography— 
e.g. iva with a fut. ind. instead of an aor. subj.—and must be examined on 
syntactical lines before we can confidently pronounce for itacism. 
Against the assumption that MSS commonly confused indie. and subj. 
may be set the round declaration (Blass-Debrunner 17) that in NT there 
are only a few traces of quantity-levelling. Rudberg! remarks that “ D 
observes quantity-distinctions very well, and distinguishes indic. and subj.” 


5. 2 


Ω § 35. In earlier Greek o differed from o as ἡ from 


εἰ it was the open 6 to which we come nearest in the 
syllable ore, as to the open @ in ere (there). In Egypt becomes much 
closer in 11/B.c. : Mayser counts from that century 50 instances of o for a, 
and 20 more when accented, 37 and 33 respectively of for o. 

We find ov for (c) quite infrequently in 
Ptolemaic papyri: see Mayser 99f., 138, where they 
are all explained as syntactical or orthographical confusions signifying 
nothing. Later exx. such as ἀγωνιοῦμεν, διαιτουμένων, ἀγαποῦντες, 
ἀπαντούντων (see CR xvili. 1100), are late enough to be classed as due to 
mixture of classes in the Contracta. Θυρουρός in papp. and Mk 195: D*, 
Jn 108 D, is merely on the analogy of κηπουρύς etc. Ζηλοῦτε Gal 417 and 
φυσιοῦσθε 1 Co 45 are doubtless subj. (see §§ 84, 85), but owe their form 
to the assimilation of indic. and subj. in Contracta which started naturally 
from the -d class. Depending as before on his syntactical presupposi- 
tions, which will be examined in their place, Schmiedel gives the follow- 
ing well-attested reicienda: Mt 7° καταπατήσουσιν, Gal 2* καταδουλώ- 
σουσιν, Rev 64 σφάξουσιν, 930 προσκυνήσουσιν, 27% μετανοήσουσιν, Ac 7* 
δουλεύσουσιν (Gal 417 ζηλοῦτε, 1 Co 4° φυσιοῦσθε---5866 88 84-5), Mk 1550 
σταυρώσουσιν, Mt 18:9 συμφωνήσουσιν, Rey 1315 προσκυνήσουσιν, Jn 17° 
γινώσκουσιν, Tit 24 σωφρονίζουσιν, Rev 125 τρέφουσιν. He bases his 
rejection of these, as of those in ὃ 34, on the evidence of confusion of 
vowels in the Egyptian vernacular Greek. ‘ 


Q and ou. 


1 Nilicher Text 15. 


76 A GRAMMAR OF NEW TESTAMENT GREEK. [8 35 


The perplexed forms of ἀνάγαιον (so WH) are noted 
under e above: a form ἀνώγειον would account for the 
MGr, but ἀναγ. must have been the earlier—it is a choice between ἀνά 
and ἄνω for first element, and on the analogy of κατάγαιον we should 
expect the former, which is in the best MSS. WS 51 compares the 
alternative forms καταφαγᾶς and κατωφαγᾶς in Phrynichus (NP 497). 


0. Ἢ: 


land ε. 


Ω and a. 


Dissimilation takes place in ἁλεεῖς, from ἁλιεύς, for 
which WH’s Δεκελεεῖς is an illusory parallel, since 
the nom. sing. will be Aexede(e)evs.1 “Adeets is found in LX X (Thackeray 
84), and can be illustrated from P Flor i. 127!> (a.p. 256): six papyrus 
citations must be set against this one. On the contraction which more 
commonly affects two concurrent 2-sounds see below, ὃ 38. The Latin 
legio is transliterated λεγιών, but λεγεών appears not only in ACw at 
Mt 2655, Mk 515, Lk 8°, but also in papyri—see CR xv. 33), 434), 
xvili. 107b. Among 31 occurrences of λ. and deriv. in papyri I find 6 
with e; but when only i/a.D. exx. are counted e has 4and.has 7. The 
« was probably older, and marks the fact that Latin ὁ was more open than 
Greek «. A similar case is λέντιον for lintewm Jn 134, where in the 
second syllable we note the more open sound of x before a vowel. Cf. 
Lindsay Latin Language 30, and note the transliteration Ποτίολοι from 
Puteoli. The late νηφάλεος, which is not an orthographic variant, 
appears in Tit 2? D*K, and in later uncials in 1 Tim 32", 

There is no relation whatever between ἱμείρομαι 
and the rare verb ὀμείρομαι, on which see ὃ 95. 

SroBadas Mk 118 ACSVXYP (στυβ. N) is probably 
a mistake of a period in which o and v were locally 
reduced altogether to the 7-sound. It is not worth considering as a 
possible reading of the autographs. Jannaris Gr 52f. would actually 
have us believe that οὐ was monophthongal in the age of Thucydides: 
that he can quote οἴκει as a phonetic equivalent of οἴκοι is sufficient 
illustration of his equipment in comparative philology. 

The most conspicuous and universal of all itacisms 
is the complete equivalence of « and εἰ in uncials as 
in papyri and inscrr. of the Hellenistic period: whether ὁ is long or 
short it may always be replaced by εἰ. Scribes have their individual 
preferences. Thus Lake remarks (Codex Stnaiticus, p. xi) that scribe A 
of δὲ prefers 1, while scribe D prefers ε. WH App.? 159f. give an 
elaborate analysis of the orthography as determined by the best uncials. 
They note (p. 306 f.) that “δὲ shews a remarkable inclination to change 
ec into«, and B to change « into εἰ. . . the converse confusions being 
very rare in both, and particularly in B. Hence B has to be left 
virtually out of account as an authority against unclassical forms with εν, 
and & against unclassical forms with εἰ; while in the converse cases the 
value of their evidence remains unimpaired, or rather is enhanced, allow- 


| and o. 


| and οι. 


| and eu. 


1 Cf. Meisterhans? 42. 


§ 35] SOUNDS AND WRITING. rhs’ 


ance being made for the possible contingency of irregular permutations 
here and there.” Set by this Thackeray’s account (p. 86f.) of the ways of 
XB in LXX. An analysis of p!* (before a.p. 350, perhaps before 325) 
shows 12 instances of εἰ for short «, 8 for τ, while there are 16 ofc for ει, 


one of them corrected (Atrov[pyav). This summary, reinforced by the 
abundance of εἰ for (in vernacular inserr. and papyri, makes it futile to 
differentiate τ and ¢ for the uncial period. WH are struck by the good 
attestation of γεινώσκω and γείνομαι in the best NT uncials, but we may 
assume with safety that it is purely accidental. According to Mayser 
(87) the change of é (whether the old diphthong εἰ or the lengthened e) to 
7 dates in Egypt from c. 200 B.c., while in Attic, Pergamene and Magnesian 
inscrr. it is a century later. In Boeotian dialect the change was estab- 
lished in ν 8.6. (ef. Thumb Dial. 223). Wackernagel acutely points out 
that the new future δανιῶ in LXX from δανείζω convicts the translators 
and not merely the MSS of itacism.1. In agreement with this is the fact 
that in papyri of 111 B.c., except in carefully written official papers, εἰ is 
constantly changed tou, and c to εἰ (half as often), without any appreci- 
able difference between accented and unaccented syllables : the εἰ for 
v is much commoner when unaccented. Attic does not begin to show εἰ 
for i till ii/a.D., Pergamene still later, and Magnesian much the same.? 
We may safely conclude that for the NT, of which probably no part 
was written in Egypt, there was complete equivalence of εἰ and τ, but 
that we have no adequate reason to expect from the autographs εἰ for τ. 
A practical inference is that it is perfectly futile to follow our best 
uncials in printing abnormal forms like ‘ov? for εἶδον and ἱστήκειν for 
εἱστήκειν. It would be quite reasonable to accept an unaugmented aor. 
indic. in the one case, and to suppose the analogy of torn operative in 
the other. But the MS evidence is not adequate proof that such forms 
really existed. A still more practical inference is the futility of insisting 
upon εἰ δὲ for ἰδὲ in Jas 3° : see Mayor in loc. 
Before vowels, which in practice means before o 
(ΕἸ and e. 2 : : 
and a, the é-sound appears to have remained unmodi- 
fied even in i/a.D. This is shown by the very frequent spelling with ἢ 
in Attica, Asia and Egypt, especially during the Augustan period. This 
is essentially the same as the « which appears eg. in δωρεά and 
᾿Αρεοπαγίτης, Where e=older εἰ : so also Aivéas, ᾿Ανδρέας, where the e has 
the accent. This involves a sharp distinction between nouns in -edé and 
-ia, where the difference of accent persisted in nom. acc. sing. and nom. 
pl., the quality of the penultimate vowel and the quantity of the ultima 
retaining their difference also. Their coincidence in the other cases, 
together (perhaps) with Ionic influence, produced no doubt some con- 
fusion ; but it is very doubtful whether our MSS are good witnesses in 


1 Reviewing Helbing, 7iLZ, 1908, p. 637. See Thackeray 85 ff. 

* For these statements cf. G. Meyer Gr*® 180ff.; Meisterhans? 48 ff. ; 
Schweizer Perg. 52 ff. ; Nachmanson Magn. 40 f. 

3 Wackernagel (in his review of Helbing just cited) calls ἴδον an ‘‘ imbecile 
spelling.” 


78 A GRAMMAR OF NEW TESTAMENT GREEK. [8 35 


any such cases. Since the accent is the only element concerned which 
normally remains unchanged, and documents of iv/A.p. and later! are 
admittedly affected by the complete equivalence of εἰ and ὁ prevailing in 
their own day, it seems best to spell according to the classical tradition, 
not because we can prove it for the autographs, but because the auto- 
graphs are in this matter unattainable and no real difference is made. 
It may be worth while therefore to cite from the lists in Meisterhans® 
50 55 the Attic spelling of words occurring in NT. (1) (δοκιμεῖον) ; 
μείξω, ἔμειξα, μεικτός ; (σειράν) ; τείσω, ἔτεισα ; τρεισκαίδεκα. (2) ἐμπορία, 
ἐμπόριον ; ἱμάτιον ; κακοπαθία ; ὀθόνιον ; οἰκτίρω ; πτερύγιον ; σιρός ; 2 
ὠφελία. (3) Inserr. differ as to στρατεία and στρατιά for “ expedition”: 
“army” is always orparia. For the rest, it will suffice to refer to WH 
for the words in which they think unclassical spellings attested. They 
include “dors” as an alternative for ἀστεῖος, which means shifting the 
accent: here again δὲ hascin Ac 77°, Heb 1155, and its evidence may be 
disregarded. The only word that matters is -ελειπὸν (-ev) “in places 

. where the aorist would be the most natural tense”: since 8D with 
τλιπ. are faced by AC 33 al. with -λειπ., it is hard to see why the less 
appropriate aor. should be preferred. The syllable in question is un- 
accented, and the difference in sound between εἰ and. would be infini- 
tesimal. It is hard to set aside, even on small matters, judgements based 
on so unique a knowledge of the MSS. But since in this place (Heb 
1025». App.” 162) B is not extant, and ND are to be deducted from the 
list of “the better MSS” on whose “constancy ” WH depend for their 
reading, one feels blind obedience difficult. A little problem of spelling 
arises In 1 Co 35 π(ελιθοῖς : see WS 185 n?*. A ἅπ. eip. like this may be 
written indifferently, and analogies are mostly lacking—perhaps πειθός, 
as a new adj. straight from the verb-stem, is best. But Schmiedel’s 
supposition, that tre1601 codiac has had an intrusive sigma put in, has 
great plausibility : we may follow FG and omit λόγοις. On the doubling 
of o etc. at a line division see § 31: the denial of such a possibility in 
an archetype of NB etc. seems to involve a too childlike faith in Vatican 
and Sinaitic infallibility. 


fae 


Some confusion with ov appears in the word 
κολλύριον Rey 3!8 NC 046 cu%°, where APw have ov. 
See Thackeray 92 for variations in LXAX MSS. Blass ΟὟ. 22 pronounces 
the ov “certainly” due to Latin influence: Debrunner 27 drops this, 
only noting (from Crénert Mem. 130) that ov is later. The papyri vary : 


Y and ou. 


1 In the cursive period a reversion to classical spelling as such may be seen 
everywhere. 

2 This word supplies a good test of WH’s method. They declare σειροῖς 
“certain” in 2 Pet 24, Here & has σιρ. and B σειρ., and on the principles 
quoted above from WH neither of these counts. The “certainty” of cep. 
seems to depend on AC, unless the witnesses for σειραῖς are to be counted. But 
this word had εἰ in Attic. 


§ 35] SOUNDS AND WRITING. 79 


for v stands P Oxy viii. 1088? (early i/a.p,); for ov P Flor ii. 1772° 
(4... 257)—P ΒΥ] i. 29(a)** (ii/A.D.) is ambiguous. An inscr. from 
Rome apparently from Caracalla’s reign, JGS/ 966!° (=Cagnat i. 41) 
has v. The corruption of λύσαντι Rev 1° NAC into λούσαντι P 046 is 
hardly a. case in point, for a supposedly easier sense accounts for the 
v.l, better. 

Changes between v and « are practically always 
due to assimilation or dissimilation. Βήριλλος 
Rev 21° A is a negligible exception. Assimilation changed βυβλίον to 
βιβλ., and (in reverse direction) ἥμισυς to ἥμυσυς. That βυβλίον is much 
commoner than its primary accounts for the equally regular spelling 
βίβλος (in Mk 1276 and Lk 20% D has βυβλ.). Meisterhans? 28 cites 
Attic inserr. down to ii/B.c. with βιβλίον and βίβλος : from i/B.c. the 
original v comes in for a while, Kretschmer Vaseninschriften 119 f. showed 
that in Attic a short unaccented v passed into ε if a neighbouring syllable 
contained . The Attic reversion to type in later inscrr. shows that 
there βύβλος never completely yielded to levelling : according to Moeris, 
Plato said βιβλία but Demosthenes βυβλία. See Schweizer Perg. 99 f. 
In the papyri βιβλ. predominates considerably: see Vocabulary, s.v. 
Ἥμισυ was in the papyri frequently assimilated ἥμυσυ : see Mayser 100, 
who says it was overwhelmingly predominant in iii/B.c. After this the 
«came in again, by the influence of oblique cases. We find ra ἥμυσοι 1 
(with -o.=-v by itacism) in Lk 198 D*, See further § 73. The island 
Μυτιλήνη was so named in v/ii B.c. (Meisterhans® 29): from B.c. 100 it 
became progressively Μιτυλήνη by the same dissimilation which is illus- 
trated above. So Ac 20!*, except for Μυτυλίνην L, like ἥμυσυ. The next 
verse has TpwyvAia D, from which we may fairly presume Tpwyidi(ov) was 
produced by later assimilation in MSS of Strabo and NT. 

The very common confusion of ὑμεῖς and ἡμεῖς in 
the MSS goes back as far as 1i/B.c in papyri: see 
Mayser’s instances, p. 86. The phenomenon is rather perplexing, for we 
have plenty of evidence that neither ἢ nor v had found their MGr goal, 
the 7-sound, for centuries after 11.8.6... unless it was in strictly limited 
areas. Thumb, Hellen. 150, 193, thinks that approximations of v to 7 
were due to foreign influence, esp. in Asia Minor; but we can hardly 
apply this principle to Egypt in the Ptolemiac period. Is it possible to 
set down its origin to assimilation of the vowels in nom. and dat., the 
other cases following suit? The ἡ and v were in the weak position before 
the stress, and the difference between an unstressed ὁ and ἐῤ would be 
easily slurred when the next syllable had an accented 7 It was the 
confusion in pronunciation between the pronouns which ultimately pro- 
duced the new forms ἐμεῖς (from ἐμέ, but barely differing from ἡμεῖς) and 


ἐσεῖς. 


Υϑδηῃᾶι. 


Ὑμεῖς and ἡμεῖς. 


1 Robertson Gr. 199 wrongly reports D as reading ἥμυσον. The corrector 
has written H over the OI, so as to indicate τὰ ἡμίση. ““ The reading ἡμίση 
can be seen to be later by observing that the first stroke of the H is not straight, 
and the cross-bar was made from right to left, and not carried through to meet 
the other limb” (Dr. J. R. Harris, letter of Sept. 4, 1915), 


80 A GRAMMAR OF NEW TESTAMENT GREEK. [8 35-36 


“Αλυκός (Jas 312) is really a different word from ἁλικός. Moeris makes 
v Attic and « Kown. See Mayser 102. 


8. Diphthongs with ι. 


§ 36. These were all monophthongal in pronunciation before the 
beginning of the Byzantine age. When our oldest MSS were written, the 
three short diphthongs were respectively ὁ, 7, and #, the three long ones 
had lost the « element altogether. 

(a) On & most of what need be said has been 
given under ¢«, where it is shown that the uncials 
confuse the two completely, the pronunciation having 
become identical. The date at which the change took place naturally 
differed locally. In Attica it become conspicuous from a.p. 150 on: a 
little earlier a. was sometimes written 7,' as it was in Boeotian long 
before. In Egypt the itacism appears in 11, and i/B.c., but very sparingly 
and in illiterate papyri.2 Pergamum supplies only two late stones, 
badly spelt. Magnesia has one instance of a for ε as early as a.D. 50, 
and isolated cases a little later. Nachmanson gives samples of the 
abundance available from Asia Minor everywhere in later times. 
Pompeii is specially instructive, since of course a.D. 79 is the latest 
possible date. When therefore we find (Diehl, no. 10) an iambic 
ἐϊνθ]άδαι κατοικεῖ: μηδὲν εἰσειαίτω κακόμ, with a twice scanned short 
as an itacism for ε, we have adequate evidence that in Italy of 
i/a.D. the confusion was possible. Blass however denies® that the 
fusion was generally complete till iii/iv a.p. While, therefore, Mark 
might have confused a and ¢ if he wrote in Rome and used a liberty 
possible there outside cultured circles, there is no proof that books 
written in Asia Minor would show this licence. It was complete 
before our uncials, but probably the autographs showed very little 
of it. 


Au, date of fusion 
with e. 


See ἘΝ Some materials for further inductions may be 
Avan Biblical secured by studying the Biblical or quasi: Biblical 
papyrl, papyri from Oxyrhynchus, coeval with or older than 
Nand B. Only four of these have any itacisms of this kind. P Oxy 
657, a copy of Heb. covering about one-third of the Epistie and dated in 
the first half of iv/a.p., has ε 14 times among 61 places where a should 
stand, as well as κέ once for καί, which is elsewhere universal and is not 
included in the enumeration following. It has also 3 cases of a for ε. 
The close affinity between this strongly itacistic MS and B in some 
important readings may be remembered. The two Logia papyri, P Oxy 
1 and 654 (iii/a.D.), have -a twice for -e, and 6 times rightly. The 
Hermas papyrus, P Oxy 1172 (iv/a.D.), shows ε 4/10 times. Finally we 
have in 1229 πλανᾶσθαι Jas 116 (5 exx. of correct a), and in 1230 ἔλεον 


1 Meisterhans® 34. His exx. for ε are all before vowels, until nearly the end 
of ii/A.D. 

2 Mayser 107. 3 Schweizer 78. + Nachmanson 37. 

5 Cf. Kretschmer Entsiehung 7. ® Pronune. 68. 


ἢ 36] SOUNDS AND WRITING. 81 


Rev 68 (no az): both fragments are iv/A.D. The other documents have a 
133 times in all, without a single instance of itacism.! 

ean RB. In Egy eye accordingly this itacism hardly ever 
appeared in Christian nicer documents till iv/a.p. 
and most scribes avoided it even in that century, common though it was 
in non-literary writing. The extent to which δὲ and B show it becomes 
the more remarkable ; but it does not affect their date, since the Hebrews 
papyrus is securely anchored in the same period, and is strongly itacistic 
But the authority of SB as evidence between a and ε is seriously 
shaken, just as it is for εἰ and 1, when we consider the strong claim 
that Egypt has to be the home of the @-text. 

Lint.of sub ' δ may me eee ae Schmiedel’s list 

stitutions of « οὗ /otms in whic print ε for at (τοῦ all of them 

in WH. invariably): — φελόνης, Kepéa, κρεπάλη, ἐφνίδιος, 

ἐξέφνης, Λασέα, συκομορέα, pédn. It will be noticed 

that they are all isolated words in which there is no analogy to preserve 

their older form : they were likely therefore to take on a new spelling 

earlier than words protected by association with a system. We have seen 
(p. 80) that even here the ε has a very precarious footing before iv/a.p. 

A note may be added as to relations of a and a. 
The archaic αἰεί is not uncommon in post-Ptolemaic 
papyri, but ἀεί predominates : the word itself survives mainly in formulae 
—see Vocab. s.v. There is no trace of this or of αἰετός in NT MSS; nor 
of ἐλάα (etec.), kaw, kAdw (unless ἔκλααν for ἔκλαιον in Rev δέ δὲ is to be 
counted), which Mayser 104f. notes as occurring sporadically in papyri. 
He notes that early papyri not seldom show ἴΑγυπτος (-ιστεί), under 
influence of Egyptian, as also final -a for -αι (unaccented) and in καί, even 
before consonants. If instances of this phenomenon were found in NT 
MSS, it might perhaps be taken as ἃ sign—valeat quantum !—of Egyptian 
influence. But it does not occur in δὲ, nor in 9915, in the word Αἴγυπτος 
(-os): to assert the negative for the other matter would involve an 
endless search, even if the silence of Ti could be implicitly taken as 
evidence. Ka occurs seven times in W, and in L at Mk 118. 

A special case is the passage of aw into ai, as in 
the names ’Ayaixés and Πτολεμαΐς, both of them in 
accord with normal papyrus spelling. Phrynichus considered ᾿Αλκαιΐκός 
Attic : see Rutherford NP 111-3. 

(b) Ec in its relation tox has been already discussed. 

Ev and ε (n). Some instances of « becoming ε remain to be noted. 
In a LXX quotation, Rom 913, we find Ἰχρε θ σαν in NAB*D*G, 


Αι and a. 


Aut and at. 


1 The following are their numbers and volume Terenas All: are yi D., 
except those marked* (iii/) and + (iii/iv). 

(1) LXX and OT Apocrypha: iv. 656*, vii. 1007*, 1010, viii. 1075*, 
ix. 1166*, 1167, 1168. 

(2) NT: ii. 2081, 209, vi. 847, vii. 1008, 1009, viii. 1078, 1079+, 1080, 
rey AULD, oe. 19981: 

(3) Hermas, Apocryphal Gospels and Acts: iii. 404+, 406*, iv. 655*, v. 840, 
vi. 849, 850, x. 1224, 1225, 1226+. 

VOL. 11:—6 


82 A GRAMMAR OF NEW TESTAMENT GREEK. [8 36 


against ἀχρεῖος. This is attested by a i/A.p. Cilician inser. : see Vocab. on 
ἀχρεῖος and ἀχρειόω. Meisterhans* 47f. argues that the very frequent 
spelling -na for -ea etc. (inscrr.—also papyri, esp. the Alexandrian 
collection of the reign of Augustus) proves the εἰ to have been still an 
e-sound before vowels. See exx. in Schweizer Perg. 55 ff., Mayser 67 f. : 
ef. Brugmann-Thumb 56. In 7b. 247 the history of πλέον (πλεονάζω, 
πλεονέκτης) is described : the case is not parallel with those which arose 
in the Kown period. Levelling has naturally brought in πλεῖον in most 
places (18/21 according to WH App.? 158), but not in derivatives. 
Mayser 69 shows how the εἰ form progressively ousted the ε in B.c. 
papyri: it is clear therefore that our uncials here represent their originals. 
In the case of ᾿Αρεοπαγίτης as compared with ”Apevos Πάγος we cannot be 
quite certain whether we are dealing with a Hellenistic or an older sound- 
change.! When Attic δωρειά (so till B.c. 403) became δωρεά (which was 
inherited by Hellenistic) it was a genuine loss of ὁ between vowels, and 
so when τέλειος became τέλεος (as in Heb 10! De and 3 Mace A bis— 
Thackeray 82); but when the late noun λογεία becomes Aoyéa in 
Ptolemaic papyri (Mayser 67), it means that εἰ retains an e sound before 
a vowel. The relations of ἔσω and εἰς belong to the earlier history of 
Greek. In the Κοινή, εἴσω did not survive, nor és, but we have ἔσοπτρον. 

Hellenistic shows no trace of the double value of 
ει, the ancient diphthong (written EI in early 
alphabets) and the lengthened close e (written E) : they 
had been completely fused before the Κοινή was born. One consequence 
remains, the infin. in -odv from verbs in -όω, the regular contraction of 
o+é: similarly -ἂν from -άω (not -av). The incorrect -otv—found in our 
uncials (see § 84), but certainly not in the autographs—was not derived 
from a fresh fusion of o+-ev, since for that a re-forming of the uncon- 
tracted -decy would be needed : it was only that in all other -w verbs the 
pres. inf. differed from 3 sg. pres. indic. by adding -v. 

(ὁ) Ov had the same relation to v that a had to e 
and εἰ tov: as in those cases, the approximation 
began with the long simple vowel (η, ¢ v), and then the levelling of 
quantity brought in the short vowel. Apart from its early completion 
in Boeotian, this change appeared first in Egypt, before other Kowy-speaking 
districts, as was the case with the change in a. There are sporadic 
exx. in badly written papyri of ii/B.c.; and the fusion becomes evident 
after i/A.D. Meanwhile in Attica exx. are not found till a.p. 238-44; 
in Pergamum the common people of 1i/A.D. seem to have lost the 
distinction which—as elsewhere — cultured men kept up generations 
longer ; while in Magnesia no instances are forthcoming. Schweizer 
Pery. 80 and Nachmanson Magn. 44f. give a number of exx. from Asia 
Minor generally, and by the time our uncials were written οἱ and v must 
have been everywhere identical except in rather artificial speech. 
Among the Oxyrhynchus literary documents examined above (§ 36 (a)) 


Ec in infin. of 
Contracta. 


Ou and uv. 


1 Tacitus (Ann. ii. 55, Areo indicio) transliterates with e in the primary, but 
this proves nothing for the Greek form before him. 


§ 36] SOUNDS AND WRITING. 83 


there are no exx. at all except in 018 (P Oxy 657), where we find ἐν]νυῶν 
and ἐτοιμ[πα]νίσθησαν (Heb 417, 1155), It is not uncommon in the LXX 
uncials, especially in words where the papyri soonest and most freely 
show the itacism: ἀνοίγω is so marked an example that Radermacher 
proposed to recognise the analogy of avitw—see Vocabulary, s.v. So 
λοιμαίνομαι six times in B: Thackeray (94) thinks λοιμός responsible. 
See his LXX evidence. The verb ἀνοίγω will supply a fair sample of 
the prevalence of this itacism in NT uncials:—Mt 2038. », Mk 734 Ὁ, 
Lk 1236 D, 2432 NBDA, Jn 917 DE, 26 NDM, 59,32 D, 1187 B*D, Ac 756 B*, 
98 ΝᾺ, 1210 NB*D, Rev 37 bis δὲ, 612 N8*. In W the only instances of this 
itacism are forms of ἀνοίγω in Mt 77, Mk 7**%, Lk 119 (cf. 1° avnx0.) ; 
and ov for σοί in Mk 134, Jn 24, σοί for σύ Jn 137. When we note that 
W simply swarms with the a and εἰ itacisms, and that ἀνοίγω is a very 
common word and more affected by this itacism than any other word in 
LXX, inscrr. and papyri, we have evidence enough that the o:-v confusion 
was much less conspicuous than either of the others, and still less likely 
to have figured in the autographs. One remarkable v./. should be 
mentioned as involving the itacism, viz. Ἑτοιμᾶς for Ἐλύμας, Ac 138 D¥*d 
(Lucifer and Ambrosiaster): see J. Rendel Harris, Exp. vi. v. 189-95 
(1902). The loss of « from o before vowels (not the ὁ vowels) is very 
conspicuous in early papyri: see Mayser’s tables, 
108f. It is also frequent in Pergamum, but does 
not appear in Magnesia. It was very common in Attic. Its origin 
belongs to the time when o was still a diphthong: it is significant 
that Thackeray can note only one LXX instance, ποῆσε Jer 3975 KN. 
Hellenistic sometimes uses οὐ forms where Attic had dropped the ε: 
thus ποία (grass) occurs in LXX 2/3 (Thackeray 93),! as in the Doric 
inscr. from the Asclepieum, Syll. 803!*1 (iii/B.c.)—P Lille 5° (B.c. 260) 
has πωολογ[ίαν. This substitution of ὦ for an older οἱ is seen also in 
δῴη (LXX and NT), in -ῷην from -éw verbs (“as early as Epicurus,” says 
Radermacher 73), and in πατρολώας and μητρολώας ; per contra dois, dot, γνοῖ 
and διδοῖ for the subj. in -@(s)—ef. ἔγνοιν = των Lk 16* D(al.—see Cronert 
Mem. 217), ᾿Αριστάρχοι dat. P Lille 1729 (iii/B.c.) and a few other exx. in 
Mayser 137. In all these cases » would be written, but » and ὦ were 
identical. It would be possible to explain roa, δῷη and πατρολώας on 
the same lines as cases of εἰ > ἡ before vowels ((b) above), assuming that 
before vowels οὐ was arrested in its passage towards οὐ, in the @ stage, 
where ὦ might be a graphic equivalent. The opposite change, δοῖ etc., is 
explicable always without phonetic considerations. 

Mayser 110 shows that ow tended towards οἱ, as 
ai to a (above (a)): in NT we have however 
ἀγαθοποιΐα (A -eia) and εὐποιΐα (AC -εία). 

Before leaving az, εἰ, o. we should note that there 
are words in which the use of the diaeresis, or the 
substitution of εἰ for « in the second part of the 
diphthong, shows that the two elements were still pronounced with their 


Or and o (w). 


Ou and οι. 


Diphthongs and 
the diaeresis. 


1 And quite conceivably in Jas 414, 


84 A GRAMMAR OF NEW TESTAMENT GREEK. [8 36 


own original value. ᾿Αχαΐα is transliterated Achava in Latin, and ᾽Α χαιίη 
in Herodotus shows that we have a tetrasyllabic word. The Latin 
praenomen Gaius, which was never spelt with ae, is found in vulgar 
Greek inserr. with v for 7, at a date when v was a pure ὁ sound. While 
our uncials are by no means systematic in the use of diaeresis, their 
evidence is often important. Blass, for example (Gr. 17), notes that 
Ναιμάν and Ἰεσσαί have no diaeresis in MSS, and Νεμάν is therefore not 
surprising : Naty and Κάϊν always in NB: *Hoaias 10/19 in δὲ, but a in 
B except for four places ; Βηθσαϊδά(ν) ter in δὲ, but also a ter, while B 
usually has ae. 

(d) The long diphthongs, which we write a, 7, ῳ, 
have been briefly alluded to in § 27. The “silent.” 
(« ἀνεκφώνητον) is not regularly “subscript” till 
xii/A.D., though Blass (Pronwnciation 50) notes a vii/A.D. papyrus where 
it is written a little higher or lower than the vowel to which it is 
attached. In the middle Hellenistic period it has become a mere matter 
of orthography ; and it is only because our spelling normally conforms to 
the Attic that we trouble about it at all. It is omitted in the uncials 
till vii/A.D., and ranks accordingly with accents and punctuation as a 
device of Byzantine and mediaeval orthography, useful for distinguishing 
nom. and dat. sg. fem. etc. It preserves, like the accents, a feature of 
Attic pronunciation, which—unlike the accents—had been for centuries 
extinct. 


Long ι diph- 
thongs. 


(a) Classical usage decides for κάτα (-Ξ καὶ εἶτα), but κἀγώ etc., where 
the second element has no ¢; ἦρον impf., but ἦρα aor. ete., in 
verbs where ὁ appears in present stem because of the suffix -y ; 
ζῆν, τιμᾶν etc., where our rejection of δηλοῖν (see § 36 (b)) forbids 
our altering the classical spelling : πρᾶος, πρωΐ, ζῷον, ὑπερῷον, 
Ἡρῴδης, πρῷρα, Σαμοθράκη, πατρῷος, ἀποθνήσκω, σῴζω and 
σέσῳσμαι, but ἔσωσα, etc., Τρῳάς, adv, ἀθῷος, μιμνήσκειν. See 
reff. in WS 41. 

(8) Some forms which look like datives, but were in part primitive 
instrumentals, varied in their acceptance of the -: on their 
historical basis see Brugmann-Thumb 269. The Attic inscrip- 
tions are the only safe basis, for we cannot always dogmatise as 
to the datival or instrumental origin of the form. It seems 
best to write κρυφῆ, λάθρα, πάντη, εἰκῆ, but to allow the -ἰ in 
ἄλλῃ, πανταχῆ, ἰδίᾳ, δημοσίᾳ, ete. 


Mayser (132 f., 125) shows that the omission of «in 
dative sing. is hardly found in iii/B.c. : in 11.8.0. a: ὦ 
in dat. is as five to one, though medial and initial it is 
only two to one. Against 250 exx. of correct me in 
iii/B.c. he finds only 2 of initial « and 11 of final: dropped. But while 
in ii/B.c. 414 datives and conjunctives in -y are noted, to 121 with -y, in 
a series of documents dated c. 100 B.c. (P Grenf) there are 27 of each. 
This prepares the way for the period in which ὁ is added or dropped 
indifferently, having ceased to affect the pronunciation. The disappear- 


Dropping of t 
(** subscript ”)— 
its date. 


§ 36-37] SOUNDS AND WRITING. 85 


ance of the « from the long diphthong a presents the same chronology 
(Mayser 120f). The analysis of P Tebt i. (ii/i B.c.) is instructive :— 
“The proportion of -a to -ἃ final is in official documents [including 
petitions] (nos. 5-54) 25: 14; in private letters and the land 
survey (nos. 55-103) 8:2; in contracts (nos. 104-111) 8:3; in 
private accounts (nos. 112-123) 0: 11.” 

The next stage is the irrational addition of -1, sporadically found 
even in iii/B.c., and visible in 111 (once or twice even with -a), but not 
common. 

These facts obviously show that before the second 
period of the Κοινή developed (ὁ. a.pD.—see ὃ 30) 
these long t-diphthongs had lost the glide. That this was not a peculiarity 
of Egyptian pronunciation may be seen, among other evidences, from 
Latin transliterations. Blass (Pronunciation 50) cites Thraex etc. from 
Cicero, later Thracia; citharoedus, comoedia, tragoedia, later ode and 
melodia. He also quotes Strabo (7 A.D. 24), who tells us that “many 
write datives without the cs, and reject the whole custom [of adding -c] as 
having no reason grounded on nature” (xiv. p. 648). 

It follows that in ambiguous passages of the NT 
we are as much at liberty to insert or drop ε subser. as 
to alter accents or punctuation, since none of these 
can have been in the autographs. Early translations and patristic 
quotations, if unanimous, will attest a traditional interpretation which 
must not be ignored, but need not wholly bar our freedom. 

Thus in Jn 5? Milligan and Moulton (Comm. in loc.) proposed to read 
κολυμβήθρᾳ (“. .. by the sheep pool the (pool) that is called. . .”).} 
In Heb 111} WH mg read αὐτῇ Σάρρᾳ, thus making Abraham the subject 
of ἔλαβεν. An instance of another kind is the convincing emendation 
ὃ ἂν (so D) ὠφελήθης for ὠφεληθῇς in Mk 711}; see Goodspeed in HapT 
xx. 471f. 

In forms which are born in the Κοινή our orthography is necessarily 
guided only by analogy, there being no authority unless the word 
happens to occur in well-written papyri of iii/B.c., when the « still 
represents a real sound. Whether Cronert’s instances of dam (Mem. 215) 
are early enough for the purpose is doubtful : here however the classical 
τιμῴην encourages us. But πατρολῴας (μητρολ.) has no real reason behind 
it, nor m@odoyia in P Lille 5° (δ 36 (c)). 


Latin Evidence. 


Application in 


9. Diphthongs with v. 


§ 37. Evand dv retain their full diphthongal char- 
acter, and the v necessarily resisted the tendency 
towards ὦ. Hence the spelling cov, found even in 1), at 2 Ti 41: 
(χαλκεούς), as Debrunner notes, with ref. to Cronert Mem. 128 f. Later 


Eu, Au. 


1 So Moffatt renders, ‘‘ there is a bath beside the sheep-pool.” The reading 
of W, τῇ ἐπιλεγομένῃ, points the same way ; and there is new evidence that it 
was so read in v/A.D. (#), from a Christian amulet P Oxy viii. 11517", which 
appeals to ὁ Os τῆς προβατικῆς κολυμβήθρας. 


86 A GRAMMAR OF NEW TESTAMENT GREEK. [8 37 


we find even «8; and as an ultimate result there is the pronunciation ev, 
av, which gives us MGr aftés (αὐτός). So Hebrew "Ὁ becomes Λευεί, “1 


Δαυείδ : on the later use of β in Λεββαῖος, Δαβίδ, see § 45. Οὐ on the 
other hand became completely monophthongal. See Mayser 114 and reff. 

Ev has a peculiar tendency towards av in the derivatives of ἔρευνα. 
The av forms do not appear in Ptolemaic times (Mayser 113). The 
earliest known ex. is one supplied by Wackernagel 1—e£epavynoopévous in 
a psephism from Syros, one of the Cyclades, dated mid 1.8.0. Then 
follow instances from the papyri, beginning with ἠραύνηται in P Oxy ii. 
2949.1 (A.D, 22): see Vocab. s.v. €pavvdw. Only in CP Herm 10218 does 
ἐρευνῖ . .. appear in Roman age papyri.? Crdnert Mem. 128 gives 
several instances from MSS of Philo and Josephus. Thumb Hellen. 176f. 
cites two or three papyrus exx. of ε broadening to a, which he thinks 
might be due to native Egyptian (not Alexandrian) influence. But there 
are closer parallels from Greek dialects, in Delphian ἐλαυθέραν, and 
Theran Doric compounds of εὐ- as av-.2 In LXX Thackeray 79 cites 
κολοκαύει 1 Es 43! B and πέταυρον Prov 918 BabNe.s, (The converse 
ἐντεῦθα 1 Es 5° A is clearly a mere confusion with ἐντεῦθεν.) These, 
so far as they go, might be taken as showing that there was some 
tendency towards such a pronunciation in Egypt. But in uncials of the 
NT we find it 7/8 in δὲ, 6/7 in B*, 4/7 in A, 2/4 inC, 1/1 in T, 1/2 in W, and 
0/5 in D and D,. It is rather difficult to reconcile such wide attestation 
with any very close association with Egypt, which is also discouraged by 
Wackernagel’s Syros inser.; but it may be noteworthy that the 6-text 
authorities have no trace of it—unless the fact that in Jn 75? W has a 
reading shared only with 1) latvt is to mark that passage as of δ character, 
and so neutralise the association of W with AB*T in ἐραυν. there. 
Schmiedel’s observation (WS 51) that B prefers ev in OT and av in NT 
would suit the evidence of papyri very neatly ; but it needs checking by 
Thackeray’s figures (p. 79n), whence it appears that in OT B has 13 of 
each. Proportionally therefore the statement is true. 

It is noteworthy however that it is only in the imperfect that there is 
authority worth counting for the nv in the conjugation of εὑρίσκω. The 
moods in the aorist would form a check upon a spelling in indic. that 
represented no practical difference of sound. That nv was still thor- 
oughly diphthongal is well seen by a tendency, esp. in δὲ and A, to write 
nu: moreover δὲ can even divide n|i between two lines—see § 27. The 
foreign word Μωυσῆς was trisyllabic, as the Latin transliteration helps 
to show. 

Thackeray 97 gives four LXX passages where πρεσ- 
Burns=senex is written for mpeoBevtns=legatus, and 
recalls Lightfoot’s note on Phm®, in which πρεσβύτης was rendered 
“ambassador.” The evidence Lightfoot gives for believing in a confusion 


Eu and vu. 


- 


1 See Nachmanson in Berl. Phil. Woch., 1911, 1184: the ref. is 7@ x11. v. 
65372, 

2 Undated, but apparently Roman. 

3 Nachmanson, Hranos xi. (1912) 220 ff. ; Kretschmer, DZZ 1901, 1049. 


§ 37] SOUNDS AND WRITING. 87 


consists of MS mistakes, dating necessarily centuries after our period. 
Both words can be freely quoted from vernacular sources in their distinct 
forms: on P Oxy vi. 933 (ii/a.D.), a letter to a πρεσβευτής, the edd. remark 
that an error for πρεσβύτῃ is unlikely. Thayer, however (Grimm, s.v. 
πρεσβύτης), cites πρεσβευτέροις from an inscr. from the great theatre at 
Ephesus, which may be set with Lightfoot’s MS citations. Hort (WH, 
App.” 136) prefers to read πρεσβευτής in Phm®, rather than take πρεσβύτης 
in that sense. Whether πρεσβύτης in its proper meaning is not tenable 
after all is quite open to debate: the unanimity of the great Cambridge 
triumvirate did not convince a bare majority of the Revisers, as appears 
from their unpublished first draft (“an aged man,” with margin “an 
ambassador”). 

. The long diphthongs, hysterogenous in Greek, 
(see Brugmann-Thumb 64), had a very limited place 
in the Κοινή. Ionic used wv, but it did not penetrate Hellenistic, except 
in the Egyptian names Θῶυθ and Mavons, where Thackeray 163 recognises 
an attempt to represent a Coptic pronunciation : in both words the ov 
passed later into w. So also av is distinguished from 
av only by a rather doubtful phonetic principle (see 
below). Hv is found in the augment syllable of a few 
verbs, as αὐξάνω, εὑρίσκω. There are hardly any wv.ll. 
when nv comes from av—Ac 125 εὔξανε D*, ἤξανεν P—but from εὑρίσκω 
many exx. of εὗρον etc. show that the shortening here was merely levelling 
from the present stem. 


Qu. 


Av. 


Hu, 


The papyri show very largely the reduction of av 
to a, scantily till 1.8.0. (Mayser 114), but freely later, 
especially in less educated writing. It appears often in Attic inscr. after 
B.C. 74 (Meisterhans* 154) in the case of ἑαυτοῦ and αὑτοῦ. Radermacher 
37 makes the length of the diphthong essential, in which case the common 
ἀτός must be analogical;+ but Mayser’s early citations (Γλακίου, and 
similarly Πολυδέκηι: cf. ἐχαριστεῖ In a papyrus) do not encourage the 
limitation. In NT we have *Ayovorov Lk 2: ΟΥ̓Δ, but it is probably 
Latin which accounts for this: cf. Ital. agosto. (See Vocab. s.v.) A more 
certain instance is κλαθμός, six times in W (once also in L, and regularly 
in E, according to Wetstein: see Sanders The Washington Manuscript 
21n). It is noteworthy that W never has ἑατοῦ or the like. There is 
very little of this phenomenon in LXX: see Thackeray 79, where one or 
two apparent instances are (perhaps needlessly) explained away. It was 
at first strictly limited in range. As in MGr αὐτός and (a)ros exist side 
by side(Thumb Handb. 85), we may assume dialectic difference within the 
Kown. As noted in Prol. 47, ἀκαταπάστους 2 Pet 214 AB and ἀχμηρῷ 
2 Pet 119. A (see Mayor, Comm. in loc.) are probably cases in point. They 
would be excluded if we allowed only av to be thus affected. 

Ov had become a simple # sound in the classical 
period ; as in the case of e, there was no distinction 
between the inherited diphthong and the close o lengthened by com- 


Au and a. 


Ou, 


1 So also Thumb in Brugmann 4 64, 


88 A GRAMMAR OF NEW TESTAMENT GREEK. [8 37 


pensation or contraction. As the quantity distinction weakened, ov could 
represent even ἤ, as it had done in Boeotian long before (e.g. τιούχα -- Att. 
τὔχη). It is also used regularly to express consonantal uw in Latin, as 
Quartus Κουάρτος, Silwanus Σιλουαν ὁς, Or vae ovat. 

Passing by analogy formations like d@vac! Mt 261° δὲ (often in papp., 
after γνῶναι), we note some relations between ov and ev which descend 
partly from classical dialects. The Ionic contraction ev from eo survives 
in the proper name Θευδᾶς (=Ceddwpos or Θεύδοτος) : see Mayser 114, 
148, also 10, where he gives several other names in @ev- against the usual 
Θεο-. lonic forms in such cases would start with individuals hailing 
from lJonic country. An interesting question is raised by the trans- 
literation of Lucius by Λεύκιος in P Tebt i. 33° (B.c. 112) al.: contrast 
Λούκιος Ac 131, Rom 1621, Nachmanson 61 gives several other exx. of 
Λεύκιος, and thinks a genuine Greek name (derived from λευκός) has 
affected the spelling. The fact that this common name appears with 
both these forms has naturally suggested the possibility of regarding 
Λουκᾶς as= Lucius rather than as Lucanus; but there does not seem to 
be any strong reason. <A link between ov and @ is 
implied by the frequently asserted identity of Κλεύπας 
(for KXeorarpos) and Κλωπᾶς, since if the eo did contract it would 
normally make ov, as in Θουκυδίδης etc. The contraction is very abnormal, 
and can be made plausible only by the difficulty of otherwise explaining 
the name. Confusion of ov and » in verb terminations is another matter : 
it may be questioned if we are ever to credit this to itacism. See above, 
under ὦ. 


Ou and ow. 


O and ov are not infrequently confused in early 
papyri: see Mayser 116f., where the conclusion is 
drawn that in this period ὁ, ov and » were pronounced close—cf. the 
same development with ¢, εἰ, 7. A case of a different kind is the fluctua- 
tion between Διόσκουροι (as Ac 281) and Διόσκοροι, with derivatives : 
Mayser 10f. makes the latter normal in papyri for the divine name, the 
former for derived human names like Διοσκουρίδης. Here the difference 
depends simply on mixture of dialects, κόρος being Attic and κοῦρος Ionic 
(see Vocab, 8.v.). 


Ou and o. 


cae, 


This combination, not reckoned among the proper 
diphthongs because the elements could not truly 
combine, was sometimes written ver (as Mk 9! ἐληλύθυειαν W), to show 
that the vowels were really distinct. Cronert Mem. 123 ff. shows with 
a mass of exx. that w continued to be so written in the Kown, the Attic 
v being rare. Radermacher’s note (Gr. 32 n.*) that εἰδῦα εἰδύης had replaced 
the old εἰδυῖα εἰδυίας must accordingly be read in close connexion with 
the reference to “educated people” in the text to which it is appended. 


Ves 


ΔΤ & the reading is dwve [=d6vac] which Scrivener marks as a scribe’s 
error.—{ Ed. ] 


§ 37-38] SOUNDS AND WRITING. 89 


How far the Attic εἰδῦα survived in archaising speech need not be 
discussed here: there is no doubt about εἰδυῖα εἰδυίης in the vernacular 
andin NT. Similarly there is no sign of any form but υἱός, regularly 
declined: contrast Attic inscrr., where from vi/ to i/B.c. there are only 
8 instances of vi. to 67 of t. After this vi. becomes “much more frequent” 
than v (Meisterhans* 60). It should be added that Blass 10 and 
Radermacher argue that the flexion -via -vins in Hellenistic proves that 
the « has no force, or even (Blass) was “not pronounced.” But of course 
the change of η to a after p or « in Attic was a phonetic process which 
ceased to work generations before Hellenistic arose, as such a word as 
Attic κόρη shows. We have to do with inherited forms, and the presence 
or absence of the « has no concern with the appearance of the new flexion. 


Prothesis. 


§ 38. Ἔχθές (Attic: see Rutherford NP 370-2) was 
the regular Hellenistic form, not the (Ionic) χθές : 
the two appear to be related as ἐκεῖνος (Attic and Kown) and κεῖνος (Lonic 
—ib. 4). The e- is supposed to be a deictic pronominal element. On the 
other hand ἐθέλω, where the ἐ- has a different origin, is not Hellenistic : 
we have only θέλω, ἤθελον. 

In certain districts of the Κοινή prothesis is very marked before o and 
consonant, as ἰστήλη, εἰστρατιώτης, ἰστοργή. Thumb gives a table of them 
in Hellen. 145, where 25 instances are mentioned from Asia Minor, 13 of 
them being from Phrygia. He infers with reason that we should seek 
the centre of the infection in Phrygian influence, it being natural to 
suspect some foreign admixture where the language outside Asia shows 
no real sign. Where a similarly limited prothesis before “1 


Prothesis. 


impure o” 
occurs in other regions, it may, as Deissmann suggests, be traced on the 
same principle to vulgar Latin: cf. G. Meyer Gr.® 166. 


Contraction. 


The most important and almost the only new rule 
of contraction observed in the Κοινή is that by which 
two wi-sounds were made into one. Thus πιεῖν, 
ταμιεῖον, ὑγίεια Were in the NT period πίελῖν, ταμ(ε)ῖον, dy(e)ia, and this 
spelling is reflected in our best uncials normally : in the later MSS there 
is reversion to the classical type. An independent anticipation of this 
contraction is found in Ionic w6At from -u. ᾿Αφεῖς in Rev 27° is probably 
for adies (see § 86): cf. dvaot Lk 23° 8.1 ᾿Επιεικείᾳ loses « in Ac 243 
B*, but elsewhere it and the adj. retain it. Ταμιεῖον is found only in 
L33 at Mt 2426, But πιεῖν (καταπ.) appears in Mt 20? (exe. W), Mk 1038 
(exc, 1), Αἰ 2377) (exc. B*), Rom 1427 (exe, D*), 1 )Pet 58 (exc. &*); 


Contraction. 
Two 7-sounds. 


1 Prol. 45, where the parallel διασεῖν (P Leid G!*—i/sB.c.) must be cancelled, 
as Wilcken reads διασίειν there. It is too early for that contraction. ᾿Ανασεῖς. 
occurs in Wessely Zauberpap. p. 116, 


90 A GRAMMAR OF NEW TESTAMENT GREEK. [ὃ 38 


Rev 16° (exc. AC), as well as in Mt 978: ABN, Jn 47-10 A, 1 Co 99 A, 107 
AC. (WH rather unaccountably edit πεῖν in Ac 23", following B* 
alone, but obey D* rather than 8B when they go with the multitude in 
Rom 1451, Since they have no objection to πιεῖν in 5/14 places, it is hard 
to deduce any principle.) ‘The fluctuations of the uncials are reflected in 
vernacular documents. Mayser (92) cites one papyrus of B.c. 111 for 
contraction (Σουχίωι, P Tebt i. 114!°—not a striking instance), and one 
pre-Christian inser., OGIS 194°? (8.6. 37), while vec is common to the end 
of the Ptolemaic age. Attic began to contract as early as B.c. 100 
(Meisterhans* 49f.). In Asia Minor it becomes abundant in the Roman 
period, but the date of the change is hard to fix: see instances in 
Schweizer Perg. 101, Nachmanson Magn. 69f. In Egypt we note [ὑ]γῆ ΞΞ- 
ὑγιῆ BGU iii. 912) (A.D, 33), ἀποκλεῖν τε -είειν P Oxy ii. 26514 (A.D. 81/95) ; 
but ταμιεῖον once in A.D. 199, ἁλιεῖς twice in ii1/A.D. Thackeray (63) gives 
ταμεῖον CPR 1% (4.p, 83/4) as the earliest instance ; and we put beside 
it rawe[tov in BGU iv. 1194 (B.c. 27)—the contraction falls anyhow 
somewhere between these limits. Ὑγεῖα begins in the papyri early in 
ii/A.D., and so does πεῖν (πιεῖν appearing twice in i/a.D.). Thackeray’s 
analysis shows that in LXX the BN text does not represent the spelling 
of the age when the LXX was written. It cannot be proved to agree 
even with NT autographs, as we cannot be sure of it till the end of i/a.p. 
The change would naturally affect some words sooner than others. 
Ταμιεῖον was not so obviously connected with ταμίας as πιεῖν was with 
mov ; and the softening of the y in ὑγίεια would hasten the contraction 
of a word decidedly awkward in pronunciation. The MSS record for 
such words in NT probably reflects a state of inconsistency in the auto- 
graphs. It is perhaps not without significance that πιεῖν is best attested 
in books written early, and πεῖν most certain in Jn, traditionally 
published at the end of i/a.p. If we follow the lead of the uncials, and 
edit ταμεῖον everywhere, πιεῖν in Mt, Mk, Ac, 1 Pet, πεῖν in Jn and Rev, 
we shall probably come as near the original as we can hope to do. Paul 
we must leave to be inconsistent, with πιεῖν in Rom but πεῖν in 1 Co: 
if the original copies had these, it only meant that Tertius and the 
amanuensis who wrote 1 Co differed in their practice at a period when 
πεῖν was just coming in. We must add a caveat against Radermacher’s 
account (Gr. 36) of the genesis of this contraction. He would get it from 
the spirantising of the ει, just as σιωπᾶν became σωπᾶν sometimes, or 
σαρδιόνυξ in Rev 217° A became σαρδόνυξ. It is simply a contraction of 
identical vowels coming together, just as in Ionic πόλι produced πόλι 
generations earlier.? 

The contractions which operated in classical Greek, 
and especially in Attic, were no longer demanded by 
phonetic necessity in the Κοινή, and only remained in 
being because they were inherited. It was therefore 
possible for levelling of flexion to bring back uncontracted forms like 
éppeev P Oxy vi. 8605: (Acts of John—iv/a.D.), ἐδέετο Lk 8°8 (all except 


Reversion to 
uncontracted 
forms. 


1 Cf. Nachmanson in Berl, Phil. Woch. 1911, 1183, 


§ 38] SOUNDS AND WRITING. 91 


NaBLX33). Phrynichus and Thomas regarded these as Ionic, but it is 
more likely to be a new operation of analogy—é)vop : ἔλυε : : ἔρρεον : 
ἔρρεε etc. The forms of yéw present problems which will be discussed 
in their place (§ 95): here it is enough to say that while κατέχεεν Mk 148 
is an aorist, ἐκχέετε Rev 161 and συνέχεον Ac 2177 need not be detached 
from the present stem unless syntactical considerations appear weighty. 
For further parallels οἵ, Prol. 54f., and 284. Uncontracted noun forms 
such as χρυσέων and ὀστέων are more likely to be Ionic inasmuch as they 
are specially characteristic of the Eastern Κοινή : see below under noun- 
flexion, ὃ 53 Β (Ὁ). Occasional contraction of ὄγδοον to ὄγδουν in papyri 
is explicable by analogy, if uncontracted forms of other adjectives 
survived by the side of contracted. Words like véos, where the uncon- 
tracted form was traditional, showed no tendency to contract, but see 
below on the forms of veooods. An occasional form like νόῳ from νοῦς 
(BGU ii. 3855-11/ii1 A.D.) serves as a set-off for ὄγδουν. On open forms in 
3rd decl. (ὀρέων, χειλέων) see § 58 (1) (c), and on πήχεων, ὃ 59 (1): ἄν for 
ἐάν is discussed in Prol. 43 n. 

The identification of KAwmas and Κλεύπας- ΟΠ 
which see also § 44—raises some questions belonging 
properly to commentaries. Of course Κλωπᾶς could not be due to con- 
traction : eo in Hellenistic as in earlier Greek would become ov, unless 
Ionic influence made it ev—as in Θευδᾶς for Θεόδωρος or the like, Τεύφιλος 
or Θεύφιλος (papp.) for Θεόφιλος. But some analogy beyond our reach 
might account for the variant form. Κλεόπας is normally abbreviated 
for Κλεόπατρος. I should add that, while ready to leave the decision to 
the Semitists, I myself find the identification of Κλωπᾶς, KXedras and 
᾿Αλφαῖος an extremely hard saying. 

The Hellenistic indifference to the confluence of 
vowels, due to the slower pronunciation which has 
been already noted, is well seen in the levelling which 
assimilates rerpa-dpyns to other compounds of rérpa-, and made even 
ἀρχι-ιερεύς possible (Mt 261: B, P Petr iii. 53 (p)?—iii/B.c.). See Ti on 
Ac 13'. So with late cpds. like ἀγαθοεργεῖν (1 Tim 6'8—not Ac 14!7), 
against the inherited κακοῦργος and ἱερουργεῖν, and numeral forms like 
δεκαοκτώ, τεσσαρακονταετής. This feature of the Κοινή makes it very 
plain that classical scholars of the last generation were yielding to their 
besetting sin when they ruled out (eg.) etymologies of ἐπιούσιος that 
broke the laws of “correctness” by allowing hiatus. We shall see in 
this example (see § 120), and many others, that Hellenistic must be held 
as subject to its own laws alone. Analogy, and the retention of local 
dialectic forms, may account for a few new contracted forms, as -ὄγδουν 
(see § 72), ἐλᾶν Ξε ἐλαίαν P Ryl 11. 130", 2314 (Α.Ὁ. 31 and 40), φρῆτος from 
φρέαρ Letr. 12 (ii/a.D.). On the other side is veounvias Col 216 BEG, 
Ionic for νουμην.: it is not quotable from papyri or inserr. till ii/a.p.,? 
and is doubtless afl. Ἱερωσύνη for Attic ἱερεωσ. is Ionic, but need not 


Q and eo. 


New forms 
with hiatus. 


1 Cf. the compromise form ἀποχείεται, P Ry] ii. 154 (A.D. 66). 
2 Thackeray 98. 


92 A GRAMMAR OF NEW TESTAMENT GREEK. [8 38 


be contraction. Ἔν for ἄν after ὅς etc. was selected in Pro/. 42f. as a 
test of the uncials’ accuracy, and it was shown that their usage agrees 
with that of the papyri during i/ and 1i/a.D., while the use of either ἄν 
or ἐάν in this construction was very rare in the century of our oldest 
uncials. My statistics may be compared with Thackeray’s (Gr. 67f.), 
which take in later material but do not alter the result. As the 
cause of this variation cannot be phonetic, we may postpone further 
notice of it till the Syntax. 


Syncope. 


The total loss of a vowel, reducing thus the 
Syncope. number of syllables in a word, is a phenomenon 
obviously resembling the prehistoric effects of Gradation; and as we 
know a stress accent was developing in the Kowy—especially, we may 
assume, in districts where the native dialect had stress, and people had 
difficulty in pronouncing without it—we may confidently trace it to an 
identic cause. Νοσσός (-ia, -iov) is an established Hellenistic form 
exclusively found in NT: it arose from the slurring of ¢ into ἃ y sound, 
just as βορέας much earlier produced (Sopyas) βορρᾶς. The phonetic 
principle is still active, as such MGr forms as παιδία (pron. pedhyd) show. 
Attic ἑορτή kept its ground, despite Ion. ὁρτή ; but in ἐλεεινός we find 
the ε slurred once before εἰ, in Rev 317 AC (contrast 1 Co 1519, where 
only FG reads thus). Even here edeivds, with « for εἰ, may just as well 
be read: A actually has a diaeresis over the «, if I read it rightly. The 
reading ἀλλοτριεπίσκοπος 1 Pet 415 NB cu! is due to the same treatment 
of o, which became a w and dropped out, just as in the other words ε 
vanished through the stage y. Note that in all these products of allegro 
pronunciation the sound affected lies well behind the accent. Ἔσθων 
Mk 15 for ἐσθίων is not a case in point, for ἔσθω is an alternative 
conjugation. 

A different application is that of ‘“ Kretschmer’s Law,” by which e.g. 
Βερενίκη became Βερνίκη. See 88 30 and 33. Different again is καμμύειν, 
regular in LXX and NT (Thackeray 99), and warranted good Κοινή by 
the ban of Phrynichus (Ὁ) (cf. Rutherford NP 426). Forms involving 
κάτ for κατά, ἄν for ἀνά, πάρ for παρά, were abundant in poetry from the 
first, and easily if occasionally crept into popular language. It was 
forgotten that καμμύειν was a compound. 

There are not a few instances in our MSS of a peculiar syncopation 
affecting prepositions, found largely in inserr. Thus in D, Rendel Harris 
notes? κα(ταγφαγόντι Lk 15°°, πε(ρὶ) rod Ἰ. Mk 527, a(va)oras Mk 101, 
ἀ(να)πτύξας Lk 417, κα(τα)λῦσαι Ac 5°, and perhaps ἐ(πι)πλήσ(σγοντι 
Lk 2343: he observes that δὲ in Hermas is not free from such forms. 


1 Tt may be added that Jannaris Gr. 421 gives several instances of ὃς ἐάν etc. 
from MSS of classical authors. As he puts it, we should suppose Xenophon, 
Demosthenes ete. responsible for the form—quod erat absurdum ! 

2 4 Study of Codex Bezae (Cambridge, 1891), p. 147. 


ὃ. 38-39] SOUNDS AND WRITING 95 


Words which always stand in the pretone are liable to changes of this kind 
when the accent has become stress; and we see the process complete in 
MGr μέ for pera. 


Pronunciation of the Vowels, 


xe § 39. Our survey of the vowel system in 
Pronunciation detail may now be focused into a general 
and Textual _. : : : e 
Criticism,  VicW. and its results brought into line with our 
whole purpose. Pronunciation might seem 
to have but little interest for us, to whom the written page 
of the NT must be everything. But the sounds as well as 
the signs have to be studied if we would have a clear con- 
ception of the value of our oldest MS tradition. Textual 
Criticism can restore to a large extent the text of the period 
after A.D. 150, but in matters of spelling it gives us no 
evidence that is not two centuries later than that epoch. 
How far can we regard the spelling of 8B and their successors 
as faithfully representing that of the first century? Faith- 
fulness in so trifling a department of a scribe’s duty, especi- 
ally if it is maintained where contemporary usage had 
diverged from that of the first century, creates a strong 
presumption that the transcript will faithfully represent its 
exemplar in more important matters. It becomes therefore 
a really serious duty to go behind our MSS and evaluate as 
far as we can the phonetic conditions of the language in 
which Paul dictated his letters and Mark and Luke wrote 
their chronicles. Only by such inquiry can we gauge the 
accuracy of the copies on which we depend. 

We first ask as to the quantity of vowels. 
The classical language, where accent was 
purely “musical,” and there was no stress to 
lengthen or reduce a vowel, is governed by strict rules of 
quantity which are necessarily familiar to all students of 
poetry. These rules were maintained in literary tradition, so 
that a cosmopolitan collection of epigrams like the Greek 
Anthology can show us poems from the Byzantine age in 
which mistakes of quantity are almost as rare as in the 
Greek verses of a Jebb or a Headlam. But when we take 
up copies of metrical epitaphs rudely inscribed on Anatolian 
tombs in the period from which our great Biblical MSS come, 


Quantity-level- 
ling :—its date. 


94 A GRAMMAR Of NEW TESTAMENT GREEK. [8 39 


we find a great contrast. Short syllables are scanned long, 
and long as short, in odd disagreement with Homeric phrases 
and conventional terms of older poetry. When did the 
change in quantity fulfil itself, as a consequence of the change 
in accent ? 


Hatzidakis, in his monograph on the subject referred to in Prol. 34 n.?2, 
shows that quantity-levelling began outside Greece and established itself 
very gradually. The iambic quoted above (§ 36) from Pompeii shows 
that ini/a.D. both εἰ and a could be written for short cand «. Philostratus 
(ii/iii A.D.) tells us! that in his time the Cappadocians “shortened the 
long and lengthened the short vowels.” That an Athenian graduate like 
Philostratus noticed the difference tells us at once that the vowels were 
still long and short in educated Attic and that they had lost their 
quantity in the Eastern Kowy. Meanwhile in Egypt, as Mayser shows 
(p. 138 ff.), from ii/B.c. down the long vowels and the diphthongs were 
free to change with short: it is clear however that there was in Egypt 
no relation between accented syllables and length, or vice versa—see 
Mayser’s tables pp. 140f. Crossing to Roman Asia, we have in Schweizer 
Perg. 94 ff. inscriptional evidence to show that ἡ could stand for e and 1, 
that the name Νικομήδεια could scan --4 -.. and that e and ὦ were free 
to interchange : ov could represent Latin ἥ, and a and o were sometimes 
e and ὕ. The evidence is not dated as narrowly as we should like. 
Nachmanson (Magn. 63) makes the levelling begin in Asia Minor and 
Egypt in ii/B.c.: his instances from Magnesia are not very numerous, 
but are more precise in date. In his elaborate review of Radermacher,? 
the Swedish scholar refutes R.’s inference from aa in Maapkos, Ἰσαάκ etc., 
that there was a sense of long vowels needing separate notation such as 
the additional sign supplied for ὁ and 0. Attic instances of levelling are 
given by Meisterhans*, but can only be dated as “ Kaiserzeit.” It was 
natural that a phenomenon depending ultimately on the speaking of 
Greek by various peoples with native languages of their own, should 
develop earliest in foreign countries. ; 


From quantity we pass on to quality. 
Changing quality [+ has become very clear already that the 
in vowel Ε 

system. whole vowel system has taken a different 
aspect from that which it wears in the golden 

age of Athens. Any modern reconstruction by which we 
may attempt to pronounce Attic as the Athenians did will be 
almost as far from representing Hellenistic as the avowedly 
haphazard pronunciation we have hitherto used in England. 
The language is well on the way towards the pronunciation 


eg 


1 Vit. Soph. ii. 13. * Berl. Phil. Woch. 1911, Sept. 23, p. 1192. 


§ 39] SOUNDS AND WRITING. 95 


of MGr, where az and ε are 6, o and are 0, and ov u, a is 
still a, and ἡ, 4, εἰ, ot, v and 7 have all sunk into 7. But we 
have already seen that it would be a serious mistake to 
suppose that even in the age of our uncials the process was 
complete. Not even the speech of the common people had 
in any country “ etacised ” os and v, or even 7, in the century 
of the NT autographs, or for generations after. We must. 
always be ready to take our proofs from the language of the 
common people, rather than from literary documents which 
are likely to be touched with artificiality ; though of course 
we shall expect to find a considerable difference between NT 
writers due to geographical separation and to varying degrees 
of Greek culture. As knowledge of the varieties of dialect 
within the Κοινή grows, we may reasonably expect to discover 
traces in our best MSS of the minute distinctions by which 
higher criticism might argue towards a book's original home, 
or lower criticism claim some particular district for the place 
where a MS was written. 

A word should be added as to the forces 
that produced a change of pronunciation at 
least as extensive as that which marks 
modern English against that of Chaucer’s time. When it is 
recalled that the Boeotian dialect of v/B.c. was already pro- 
nouncing as as 6, εὐ as ὃ, οὐ as 1, and 7 as a close 6, it is © 
natural to conjecture a historical connexion with the same 
pronunciation in a later time all over the new Greek-speak- 
ing world. To this it is replied that Boeotian never seems 
to have affected even its next-door neighbour Attic until—on 
the assumption that this development really was due to 
Boeotian—the infection had already passed through every 
other region where Greek was spoken. It is abundantly 
clear that Attic accounts for (say) seventy per cent. of the 
Κοινή in phonology, accidence and syntax, Ionic for perhaps 
twenty per cent., and the other dialects possibly for as much 
as ten. It becomes therefore very improbable ὦ priori that 
Boeotian should affect Hellenistic pronunciation so profoundly 
without doing much more, and equally improbable that this 
very widespread charity should have failed to begin at home. 
To substitute some other single cause is beyond our power. 
In the change of the accent system we shall probably find a 


Causes of the 
change. 


96 A GRAMMAR OF NEW TESTAMENT GREEK. [S$ 39 


more powerful solvent of the vowels than any other influence 
at work. And this in its turn may well have been largely 
due to the fact that Hellenistic was everywhere (except in 
Greece itself) the language of bilingual people. Of course 
the influences of native dialects—diseussed at length by 
Thumb in chap. iv. of Hellenismus —would affect Greek in very 
different ways. But they would mostly agree in possessing 
stress rather than pitch accent; and it would be natural for 
them to impart a stress to the Greek accent. This however 
is mere conjecture. We must mostly be content to recognise 
the fact of development without attempting to explain it, 
the conditions being largely outside our range of knowledge. 


A summary of pronunciation was attempted above, 
but we must return to one problem more in detail. 
How does itacism affect questions of syntax? If (0) 

and ec were identical, and o and , was there any real difference between 
the future indicative and the first aorist subjunctive, the present indica- 
tive and subjunctive of verbs like Avw? The answer has so much effect 
on our views of verb syntax that we must examine the history of the e 
vowels more closely. We may start with the Attic m, which in ili/ii B.c. 
was predominantly changed to εἰ (Meisterhans* 38 f.), but recovered itself 
in i/.c., and afterwards dropped its «like aand ». One curious survival 
of this orthography is very general in Hellenistic (see Mayser 127 and 
the inserr. cited in Dittenberger’s index, Syll.? iii. 226 f.), viz. the perf. 
act. and mid. (not the other augmented tenses) of aipéw and cpds.! In 
᾿λῃστής and ἀποθνήσκω the older y survived, in λειτουργός the later εἰ or 
ει (perhaps through association with λιτή ‘prayer’): so also κλείς and 
κλείω. The dative τιμεῖ was replaced by τιμῇ after a short period of 
prominence by levelling action: χώρα nom. and χώρα dat. were only 
orthographically different in i/A.D., and τιμή, τιμῇ followed suit, except 
for the accent. This takes us to the relations of εἰ and ἡ, already 
described in detail. H.and εἰ have contacts especially in prevocalic 
position. Sucha table as Dittenberger’s in Syll.? iii. 226 will show how 
abundant was ἡ for εἰ before vowels. The same may be said of the 
papyri, especially during the Augustan period. H had thus two values, 
one that of an open @, as in earlier times, and the other close  : the 
former drew e to it and the latter εἰ. Ultimately ε remained an ὁ sound, 
as it does to-day ; and the old ἡ which had this value was ultimately 
spelt ε (as στήκω, MGr στέκω). The other ἡ passed at last intoz: thus 
Quirinius is transliterated Kup(e)ivos in BW, but Κυρήνιος in the rest. 
Why στήκω kept an e sound while πλῆθος took an 7 we are hardly able 
to say. 


Itacism and 
Syntax. 


1 The NT has only one occurrence, 2 Co 97, but there FG 33 show -ep: in 
Ac 5° and 2750 there is no variant upon ἀνῃρέθη and περιῃρεῖτο. 


§ 39-40] SOUNDS AND WRITING. 97 


To return to the practical problem of indicative and subjunctive. It 
pres. indic. mid., Avec and Avy, were wholly equivalent, and the indic. 
Aves -et, λύσεις -er, identical with subj. Avys -ῃ, λύσῃς τη. Restricting our- 
selves to Attic, we ask whether this identity continued. Differentiation 
could arise, as in the noun, by levelling: Aves λύητε invited assimilation 
as strongly as τιμή -ἦν τῆς -et. But the very levelling which brought back 
a new 7 tended to keep it nearer to ἡ than the old y that had passed 
into ει. Hence its readiness to lose « and attach itself rather to 7 than to 
ει. Meanwhile ἡ itself was becoming close, and in ii/a.p. had largely lost 
its difference from e even in Attica.’ Outside Attica, according to 
Brugmann-Thumb, final -7 never changed to -e. In that case there was 
an additional force tending to differentiate indic. and subj. for a time. 
There was moreover a reason for the survival of the normal -y (=-e(o)av) 
in 2 sg. pres. ind. mid.: the three exceptions βούλει (still in NT), οἴει, 
ὄψει, may perhaps be treated as Attic forms surviving through the influ- 
ence of idiomatic turns of speech in which they occurred.2, We have 
thus a reason for the survival of a separate flexion for indic. and subj. 
until the time when y and its equivalent » had sunk into 7. By that 
time ο, and ov were very near each other, and o and ὦ in many parts 
were identical. Thus the 2nd pl. was the only person unaccounted for, 
and in an unaccented syllable 6 and 7 did not differ greatly. All this 
will account well for a partial fusion of indicative and subjunctive, such 
as we shall have to note in the Syntax: syntactical forces may have co- 
operated with phonetic in producing this. But it is extremely hazardous 
to assume that these conditions applied in i/a.D., except in districts 
which had little or nothing to do with the NT, and in an educational 
stratum lower than any represented there. 


Aspiration. 


: ἣ § 40. The early uncials have breathings marked as 
Pa 10 ittle as they have punctuation or accents ; but lite- 
; rary documents have begun to insert them at a date 
not much later than that of δὲ and B. They appear in W, where 
Sanders (p. 18) says the rough breathing may be rarely seen on 
monosyllables, and on words liable to be confused: he counts “29 
instances in Mt; none in Mk 1-5*; 3 in the rest of Mk; 44 in Lk; 
and 4in Jn 513 toend. There are no mistakes in its use. In the first 
quire of Jn there are no breathings.” This MS is accordingly the 
earliest to show them. We find breathings marked in some of the 
papyri edited in Berliner Klassikertexte vi. by Schmidt and Schubart 
(1910)—a v/a.p. codex of Ignatius, and an anthology of Basil’s letters 
(same cent.) : see CA xxvii. 176f. (The sign + was used for h by gram- 
marians perhaps as early as Aristotle ; see Blass Pron. 92), 


1 See Brugmann-Thumb 62-4, Nachmanson Beitrige 37, Meisterhans? 19. 
2 But the history of this form is much disputed : see Brugmann-Thumb 405, 
8 Radermacher 37 dates this approximation in iii/A.D, 


VOL, Il.—7 


98 A GRAMMAR OF NEW TESTAMENT GREEK. [Ἃ8 40 


A point of interest arises in these documents, 
where we find adpartoi (p. 5) and an@éctHke (p. 26). 
The second of these does not represent classical pronunciation, since the 
h was absorbed in combination with the τ, while ἀόρατος certainly lost 
any aspiration centuries before this papyrus was written. But it may 
be assumed that while the aspirate lasted it was pronounced in medial 
position in such cases: cf. such Latin transliterations as Huhemerus, 
Euhodius. As h was not yet dead in i/a.D., we may safely infer that 
εὐόδοῦν and the like would represent pronunciation more accurately than 
our conventional spelling. Alexandrian grammarians wrote the “ inter- 
aspiration ” in the texts of the poets ; but there is no doubt that the h 
was even weaker in this position than it was when initial. See Blass 
Pron. 96. 


Interaspiration. 


ὶ Initial A was weak, but still pronounced in our 
sae ae period, and (as in other dialects where it is weak) it 
ity is not infrequently misplaced. There is a series of 
wrongly aspirated words which obtained such a hold that a trace even 
survives in MGr, where an initial ἢ has not been pronounced for many 
centuries. This is épéro(s) “of this year,” which takes us back to ἐφ᾽ éros, 
a very common combination in papyri and Κοινή inscriptions. So 
καθ᾽ éros etc. This form does not happen to occur in NT, for in Lk 95 
the newly discovered W alone reads καθ. But “ἐφ᾽ ἐλπίδι, accepted 
Rom 82°, has some primary authority (&?. Al. B!. Cl. D4. D,". 655) 8/9 times, 
besides ἀφελπίζοντες 1/1” (WH App.” 150). This last is read by DP in 
Lk 65, The occasional aspiration of εἶδον and its compounds, accepted 
by WH in Ph 258 and Ac 27mg, is found in the best MSS; while ov 
ὁλίγος appears 6/8 times in Ac (N*. A’. Bl. D!) asin LXX twice. Καθ᾿ ἱδίαν 
occurs “ 9/16 times (N!. BS. D3. A!”—add W!). Αὐθόπται stands in Lk 12 W, 
and οὐχ ὅψεσθε in Lk 172 A. A series of such forms is noted from Ὁ 
by Rendel Harris in his monograph on the MS1—z6vvarol, eidov®, ἴδιος, 
oriyos!, ἐφίστασθε!, ἐφαγαγεῖν", ἑμοῦϊ, ἐμέ!, ἑλπίζω!, ἑλπίς". 
The causes of this phenomenon must be sought 
Due to analogy. in the working of analogy. Thumb puts this con- 
cisely in Brugmann Gr‘. 143, where he shows how in classical Greek 
ἡμεῖς, ἧσται and ὁκτὼ évvéa (in Heraclean Doric) followed respectively 
ὑμεῖς, ἕζομαι and ἑπτά, etc. “In the Κοινή this process went further, 
e.g. ἔτος after ἡμέρα, ἵσος after ὅμοιος, ἐφιδεῖν and ἐφόπτης after ὁρῶ : see 
Thumb Spiritus Asper ΤῸ ff., Mayser 199 ἢ. Common formulae like 
καθ᾽ ἡμέραν, ἐπ᾽ ἴσῃ καὶ ὁμοίᾳ, and the unifying effect of a paradigm like 
ἐφορᾶν : ἐπιδεῖν : ἐπόψεσθαι, help us to see how the analogy worked. 
Καθ᾿ ἱδίαν may have followed καθ᾽ ἑαυτόν. “Ed” αὕριον (as P Tebt i. 
11917, B.c. 105-1; P Ryl ii. 4414, ili/a.p.) obviously follows ἐφ᾽ ἡμέραν, 
and is itself an argument fora similar account of ἐφ᾽ ἔτος. Οὐχ ὁλίγος 


1 Where (p. 138) he also notes a dropped aspirate, in εὑρίσκω, κατεξῆς", 
οὗτος! and ἑαυτῶν!. The explanations offered in this chapter (dated 1891) are 
antiquated by later developments of comparative philology. Some of the 
instances included are discounted by the special conditions affecting οὐκ and 
οὐχ : see below. 


§ 40] SOUNDS AND WRITING. 99 


arose later than the rest, being rare in LXX, where other cases are fre- 
quent : see Thackeray 126, where an explanation is attempted, based on 
a (less probable) account of the common Hellenistic ἐφιορκεῖν (Mt 5° δ), 
ἐφίορκος (1 Tim 119 DP). Brugmann-Thumb 166 explains the latter as 
a mixture of ἐπιορκ. and ἐῴφορκ., which are lento and allegro forms respec- 
tively. Since it is only with οὐχ that these aspirated forms of ὀλίγος 
occur, we cannot safely draw any inference: see below, and cf. Cronert 
Mem. 152 n.4, where one instance of μεθ᾽ ὁλίγον is the only exception. It 
is less easy to suggest analogical cause for ἑλπίς, which (Prof. Souter tells 
me) is supported by proper names in Latin inscrr.—Helpis, Helpidius, 
Helpidophorus. Was ἑλεῖν in thought ? 

Since even Thackeray 1 still clings to the old idea 
that the lost F produced this h, it should be observed 
that a large proportion of forms already mentioned—. 
add from papyri such types as μεθηνέχθη, καθ᾽ ἑνιαυτόν, 
"Ed ᾿Ασκληπιάδου, μήθ᾽ ἄλλον, ef’ ὄνοις, epayayetv—cannot be explained in 
this way. And there is this further objection, that the phenomena in 
question are late, whereas in Attic and Ionic F totally disappeared in pre- 
historic times. We should have to assume, for example, that Fedzmis 
generated «Amis in some dialect which ultimately influenced the Kown, 
but left no trace of itself until the h outcropped in post-classical times. 
For this there is nowhere any evidence; and even if (as in the case of 
é\mis) we have no complete explanation to offer, it is safer to assume 
some hidden analogy. Can we always account scientifically for our own 
coekney’s sins of commission and omission in this matter? The analogy 
is the more instructive since contemporary Rome had “’Arry” similarly 
employed : cf. Catullus 84 : 


Not connected 
with long lost 
digamma. 


Ἢ Chommoda dicebat, si quando commoda uellet 
dicere, et insidias Arrius hinsidias. 


It saves us a good deal of trouble to realise that when h is feeble it 
always tends to be misused in ways which cannot be wholly accounted 
for. 

See on the whole subject Thumb’s monograph on the Spiritus Asper 
in Greek (Triibner, 1889), also Crénert Mem. 148 ff. ; Schweizer Pery. 
116 tf. ; Radermacher Gr. 38 ; Mayser 199 ff. 

Miscellaneous problems involving the aspirate 
may be collected here (mostly from WH App.? 
150 ff.). 

Ἐπίσταται 1 Th 5° NBL 33 (Wisd 6° B) is an isolated slip, to be set 
beside ἀποκατιστάνει Mk 913 B* (where however the true read- 


Miscellaneous 
irregularities. 


1 And later still A. T. Robertson Gram. 209. It should be observed that 
he fails to represent the philologist’s case against the digamma as a vera caussa 
for the Hellenistic aspiration. No one doubts that ὦ sometimes represents F: 
it is found in three words beginning with ves-, as explained by Thumb in 
Brugmann Gr.‘ 52. But all this belongs to a period centuries earlier than 
hat in which καθ᾽ éros and the like began to appear, when F was absolutely 
dead except in a few corners where old dialects still struggled for life. 


100 A GRAMMAR OF NEW TESTAMENT GREEK. _ [§ 40 


ing may be ἀποκαταστάνει N*D—cf. MGr στάνω, a form thrice 
foundin D). The converse is found in ἐφίστασθαι for ἐπίστασθε, 
Ac 1055 D*. WH and Schmiedel take both as “thoughtless 
confusion of the verbs ἐφίστ. and ἐπίστ." (WS 39). 
Αὑτοῦ (etc.) is read by WH some twenty times: see the question dis- 
cussed below, § 76. 
Οὐχ appears in LXX and NT not infrequently where οὐκ is expected. 
Thus οὐχ ἠγάπησεν Rev 121] A, οὐχ Ἰουδαικῶς Gal 214 R*ACP 
33 (οὐχ Ἰούδα Sus δ΄ ABQ, according to WH p. 314), οὐχ 
ὀλίγος (see above), οὐχ ὄψεσθε Lk 177? A (see above), οὐχ ἰδού 
Ac 27 NDE 61; οὐχ ἐπόνεσαν, οὐχ εἰσακούσομαι, οὐχ ὠδῖνες, οὐχ 
ἰσχυρός in LXX (WS 39). Some of these cases may be specially 
explained ; thus WH would make Ἰούδας the regular represen- 
tative of 7737. But there does not seem to be strong reason for 
our seeking to take them one by one. Moreover there are 
instances of οὐκ for οὐχ, as οὐκ ἕνεκεν 2 Co 712 NCDE 33, οὐκ 
ὑπάρχει Ac 3° XC, οὐκ εὗρον Lk 245 ΟΣ (“ etc.,” says Schmiedel), 
We must apparently allow exegesis to decide between ἕστηκεν 
and the Hellenistic imperfect ἔστηκεν after οὐκ in Jn 844, 
: An important class of variant forms due to meta- 
ΡῈ οΥ thesis of aspiration came into the Κοινή through 
‘Tonic influence, and are shown by MGr to have 
maintained their position: see Prol, 38. Χιτών κιθών, χύτρα κύθρα, 
ἐνταῦθα ἐνθαῦτα, Barpaxos βάθρακος, φάτνη πάθνη, are instances, and there 
are further complications due to mixture, such as χιθών and κιτών. See 
plentiful exx. from papyri for κιθών etc. and κύθρα in Crénert Mem. 82 n.3, 
83n.!; also Mayser 16, 41 and 184, Thackeray 103. Tracesin NT are not 
many: χειθῶνας Mt 1010 D*, Lk 95 W, κιτῶνας Mk 14° B* (“ut alibi 8” 
Ti), warvns Lk 131% W. See Brugmann-Thumb 121f. Rendel Harris 
(Codex Bezae 140) points out that in Lk 31! D has χιτώνδο, with a smooth 
breathing: this is by way of correcting his exemplar, which had χιθ. 
Dr. Harris gives other instances of this metathesis of aspiration from D 
(or its Latin), viz. Ac 1616 phytonem (1.6. *pirwva), 161}: Samotrachiam 
(* Σαμοτρᾷχην), Mk 79 ἀτεθεῖτε. 

Words beginning with a single p, with another p at the head of the 
second syllable, were according to the ancient grammarians 
(see WS 40) without the usual‘. Thus ῥεραντισμένοι ete. This 
is in accord with the rule which makes reduplication involve 
loss of aspirate (τίθημι etc.), if we may assume that the voiceless 
initial » kept its quality when pushed into a second syllable. 
But it is not really certain that initial p was voiceless except 
when it came from sr-: see Brugmann-Thumb 145. 

The breathing has been variously determined for ὀμείρεσθαι (see § 95), 
εἰλικρινής and -ia (8 105), ἀλοᾶν (cf. ἀπαλοᾶν and καταλ.) and 
ἅλυσις. 

In Semitic loan-words WH use ‘ for 7 and nm,’ for κα andy. There is 
no reason for writing ὕσσωπος (372) except that initial v 


regularly has‘: apparently even English has only aspirated 


§ 40-41] SOUNDS AND WRITING. 101 


the word since the Genevan Version. There is difference 
between editors as to the breathing in the name Jerusalem. 
WH (Jntr.? 313) refuse the ‘to Ἰεροσύλυμα as coming from a 
“false association with ἱερός." But however “ false,” it may well 
be quite real as popular etymology. In ᾿Ιερουσαλήμ Blass is 
right in giving “I, since the association would not affect the 
indeclinable. 


Single and Double Consonants. 


§ 41. Ambiguous cases under this heading are 
practically restricted to the continuous sounds, the 
Liquids and Nasals and o. We find in papyri and 
inscriptions of the Hellenistic age, as in inscriptions 
of classical times, a tendency to double and a counter-tendency to drop 
one of the elements ina double: thus we have τἄλα, ἀναβάλουσιν, παρησία, 
ypaparis, yevai(ov), ἔλασον ete. in early papyri (see Mayser 212-4), and 
again διάλλογον, ἐνναντίων, εἰσσανγελέων, ἔλλαττον, “Qppov, λατομμίδα, 
ἀννανεώσεις, ἐκπεσσεῖν (2b. 217-9). Instances with mutes are virtually 


Single and 
Double 
Consonants. 


negligible. 

(a) Ἐννέα but ἔνατος ἐνενήκοντα must be kept as in earlier Greek. D 
shows évvaros, an obvious case of analogy. 

Τένημα (from ¥ γεν of γίνομαι) is a Κοινή word for ‘ vegetable produce,” 
and must be sharply distinguished from γέννημα (from γεννάω) 
“ offspring.” 

’Eveds has ν as in cl, Gr. 

Ἔκ- and συν-χύννω are supported by MGr (Prol. 45 n.?): cf. ἐκτίννω 
in BGU iii. 8968 (i1/A.D.) and 1. 28240 (2b.), and ἀποτίννω in 
P Gen 74?! (? ili/a.p.), and ἀποτιννύτω Syll. 7375! (ii/A.D.). 

*Ap(p)aBov with p and pp has about equal warrant in papyri 
(Prol. 45). 

Παρ(ρ)ησία (-ιάζομαι) has p single in some of the best uncials in nearly 
one-third of its occurrences. 

Πυρ(ρ)ός and πυρ(ρ)άζω drop an p in Mt 16? C and late uncials, Rev 64 
AP 046, 12° C 046; in these last πυρός was sometimes mistaken 
for the gen. of πῦρ. 

The proper names Μύρρα (Ac 27° B Hier) and Φύγελος have the single 
liquid better attested in inscriptions : see WS 58, and note Mipa 
in Ac 21! D. 

Πλημύρης Lk 648 DW al. is the older spelling : the doubled p is due to 
a popular etymology connecting the first part with πλήν ---866 
Boisacq, s.v. Of course it does not follow that we should edit 
the single μ, which may well be due to correction by literary 
hands in some ancestor. 

(b) The older doubling of p after the augment or in composition is 
very imperfectly carried out, and in many forms vanishes 
entirely. Words which have an established existence as com- 
pounds, with the simplex no longer in conscious contact, tend 


102 A GRAMMAR OF NEW TESTAMENT GREEK. [ὃ 41 


to keep their pp: thus χειμάρρους, ἄρρητος (but ἀναντίρητος), 
ἔρρωσθε (ῥώννυμι being obsolete), ἄρρωστος, aivoppociv. But for 
words kept in contact with a paradigm the pp is in a minority : 
διαρρήξαντες Ac 1414 (but διαρήξας Mk 1468, περιρήξαντες Ac 
16°”), ἐρριζωμένοι, ἐρρέθη(ζσαν), ἔρριπται Lk 17? (but p in other 
forms from ῥίπτω), against dpados and various forms from 
ῥάπτω, ῥήσσω, pew, ῥαπίζω, ῥαβδίζω, propa, pavtitw. The 
opposite tendency may be seen in sporadic spellings like rappov 
in 2 Co 10? DE 47. 

(c) Semitic words (among which should be included the thoroughly 
naturalised ἀρραβών above) show some degree of uncertainty in 
their spelling. Ἰωάνης is accepted by WH except in two 
places, since B has it 121/130 times. But δὲ has vy except 
in one part, which Lake has now shown not to be “ written 
by the scribe of B.”! The inscriptional evidence WH quote 
is relatively late, but it is as good evidence as the ὦ priori 
considerations in WS 57. The record of D is curiously 
divergent: it has yy in Mt, Mk and Jn at the beginning, 
but ν in Jn from 5%* to the end, and in Lk and Ac. See the 
discussion of the significance of this variation in G. Rudberg’s 
Ntlicher Text, 14.2 Ἐλισαῖος and μαμωνᾶς are clear, Ἰόπ(π)η, 
Tev(v)noapér, Bapoaf(8)as are regarded by Schmiedel as more 
doubtful, on the ground of their Semitic etymology. In 
Ac 7? Ὁ has Xapav after the Hebrew: contr. Kappa (Strabo), 
Lat. Carrhae. 

(4) KpaBarros, a word of foreign origin (said to be Macedonian), is 
spelt with single r several times in B, which thus agrees with 
Latin grabatus. ACDW 3 have rr always, as also δὲ 1/11: else- 
where δὲ reads κράβακτος, a form found elsewhere only in two 
papyri, dating respectively from the centuries before and after 
that in which δὲ was written: for the bearing of this on the 
provenance of δὲ see Lake’s introduction to the facsimile of the 
MS, p. xi. The spelling κράβατος can be quoted from Egypt 
much earlier, viz. from a “probably Ptolemaic” ostracon in 
Melanges Nicole, 184, which enables us, if we like, to associate B 
also with Egypt as far as the evidence of this word goes. But κρά- 
Barros also can be quoted from Egypt: see P Lond 1910 (a.p. 
103-17) (=11. p. 265). Κράββατος appears in Ac 515 E and in 
the mass of later MSS: it seems to be a dialectic variation in 
the Κοινή, which has left descendants in MGr dialects. Cf. my 
Kinleitung 60, and Vocabulary, s.v. 

(e) Doubled aspirates in Semitic words like Μαθθαῖος, ἐφφαθά, or the 
Phrygian ᾿Αφφία, are contributory evidence of the spirant 
pronunciation. 


1WH App.? 166: see Lake, Codes Sinaiticus xii. 
2 Upsala University publications, 1915. 
8. Ἢ has κρέβαττος twice—a form which has parallels in MGr. 


§ 41-42] SOUNDS AND WRITING. 103 


Intrusive Consonants. 


The combination op, which is not found in 
native Greek words, sometimes developed a transi- 
tion sound + in the name Ἰσραήλ and derivatives. 
So in Mt 1928, Mk 1229 in W, with D in the latter, and sometimes 
in Ac in B. In δὲ there is a δ, which appears 8/9 times in the word 
Ἰσδραηλείτης. The occurrence of Ἰστραήλ in magic papyri and in one or 
two Egyptian inserr. is noted by Lake (Cod. Sin. xi.) as nullifying the 
argument of WH! that the intrusive dental was a Latinism and therefore 
supported Western provenance for δὲ or B or both. The phonetic 
development was easy, as is shown by the fact that the root srew “ flow” 
(ῥέω) produced the original of our stream in Germanic and Στρυμών in 
Thracian, with the same thing in Albanian and Lithuanian: see 
Brugmann Grd.? 1. 827. 

aid labial Intrusive B nee p and p (as in ἄμβροτος, 
between μ and μεσημβρία) appears in the proper name Ιαμβρῆς, if 
the usual Hebrew etymology is right. So Mamre 

ΒΙΌΣ ΟΣ becomes Μαμβρῆ, and Samson Σαμψών (Heb 1133). 

Cases like the analogical λήμψομαι and the (unexplained) σφυδρόν 
are dealt with elsewhere. 


Intrusive dental 
between o and p, 


LInquids and Nasals. 


§ 42. Variation between ἃ and p appears in κλίβανος, 
for which Phrynichus (Rutherford NP 267 f.) claimed 
κρίβανος as Attic: the A form probably entered the 
Κοινή from Ionic (as Herodotus has it) or Doric (Sophron). As Lat. libwm 
and our loaf (A.S. hidf) show, the / was original, and Attic was peculiar. 
The Latin word flagellum is dissimilated to φραγέλλιον (ppayeAdow) : D 
keeps the A, as does a papyrus written apparently by a Roman, in 
Trajan’s reign—-P Lond 191" (=i. p. 265). The Hebrew Synba is trans- 
literated BeAiap,? which again may be mere dissimilation; but some 
Semitic etymology need not be excluded—see HB: s.v. In D the ν in 
λύχνος is twice written A, and in πνέω once: see J. R. Harris, Codex 
Bezae 143f., where also péyap in Lk 11 is connected with the ἃ in 
μεγάλη. The form πλέοντα (Lk 1255) is a happy accident, agreeing with 
a prehistoric ancestor of πνεύμων : see Walde, Lat Hiym. Wort. s.v. 
pulmo. So κάλῴφος Lk 64 W, λεφέλη Lk 934 W: also P Oxy ii. 2421? 
(A.D. 77) λαύλας. 


Variations of 
λ and p, 


The relations of pp and po affect the NT forms of 
θαρσέω and ἄρσην. Θάρσει -cire is used exclusively 
(Evv., Ac), while for the rest of the verb forms of θαρρέω occur 2 Co 
quinquies and Heb 13°, without variant. Αρσην (a derivative) appears 


pp and po. 


1 Intr.* 265. 
2 WS 58 cites Berial from Asc. Isai., but see Charles’s edition, p. 6, 


104 A GRAMMAR OF NEW TESTAMENT GREEK. [8 42 


unchallenged in Mk 10%, Lk 27%, 1 Co 69, 1 Tim 110; but ἄρρην has 
some warrant in Mt 19* E, Rev 12° 8 046, 7b!3 046 cupl; and in Paul, 
Rom 157 N*AC 33 (ter, but C has po 3°), Gal 8338 δὲ, where WH give pp 
as an alternative. Wackernagel (Hellenistica, Gottingen, 1907) shows 
that the sporadic appearance of Attic pp is parallel with that of rr, 
and normally comes into the Kown with specifically Attic words that 
contained it. Whereas, however, “the Attic tr was shared only with 
the dialects of Eretria, Oropus and Boeotia, numerous dialects agreed 
with it in the use of pp” (op. cit. p. 25). Ionian influence would be 
the only serious discouragement to the use of θαρρῶ and ἄρρην, so 
that mixture was easy. We might imagine, for instance, that θάρσει, 
θαρσεῖτε, “cheer up,” was a favourite phrase in Ionic: if that were so— 
it is only offered as an illustrative speculation—we could understand 
how the rest of the verb came from θαρρεῖν, while the imperative phrase 
lived only in the po form. This distinction is maintained in NT, and in 
LXX with only one or two exceptions; and it might easily arise from 
such cause as has been suggested. An interesting confirmation may be 
seen in W at Mk 1048, where θαρρῶν ἔγειρε is read instead of θάρσει, ey. 
The noun θάρρος was not really even Attic, and θάρσος accordingly stands 
alone. In the Ptolemaic papyri however we have po only (Mayser 220) : 
pp begins to appear later (cf. CR xv. 33), as it does even in Ionic territory 
on the inserr. of Pergamum (Schweizer Perg. 125). MGr shows mixture 
still : see Thumb Hellen. 77f. In NT ἄρσην is read throughout by WH 
(against NAC in Rom 157 and δὲ in Gal 355); but the papyri show great 
variation—see Vocabulary, s.v. Δέρρις occurs in Mk 19 D: Wackernagel 
(p. 13) notes that δέρσις never existed. 

Assimilation of the ν of ἐν and σύν takes place in 
composition according to the traditional spelling, 
which probably represents the pronunciation even in 
the Hellenistic age, to judge from the freedom with which assimilation 
takes place in inscriptions and papyri when proclities are linked closely 
with the following word. Mayser’s exhaustive presentation of the 
evidence from Ptolemaic papyri is summed up in tables on p. 231 (with 
which compare those of Thackeray 132-4, tending the same way). He 
shows that the words which most favour assimilation are τόν, τήν, τῶν ; 
ὅν, ἥν, ὧν ; ἄν, ἐάν, ἐν, ἐπάν, μέν, αὐτόν, τοσοῦτον, πλέον ; πλήν, νῦν. 
There is moreover a marked difference between the effect of labial and 
guttural initial. For a large number of papyri from ii1/ and ii/B.c. the 
proportion of assimilations to non-assimilations in separate words is for 
iii/ 1 : 4 before labials, 1:9 before gutturals ; for ii/ 1: 11 before labials, 
none before gutturals (80 exx.). The tables for ἐν and σύν in composi- 
tion (p. 234) give for ili/ 7:1 for assimilation before labials, 4 : 5 for ii/; 
before gutturals 4:1 and 5:6 respectively. Even here therefore the 
progressive ousting of assimilation is very marked, and the difference 
between gutturals and labials. 

In NT συγγενής (-ea) and σύγχυσις are the only forms with ovy- that 
find their way into WH’s list of “ certain and constant ἢ assimilations : they 
print with an alternative ovyxexaduppévoy and συγκυρίαν. Compounds 


Assimilation of 
ν in ἐν and σύν. 


§ 42] SOUNDS AND WRITING. 105 


of ἐν show 7 assim. to 7 non-assim. with ey-, and 26 assim. to 1 non- 
assim. with ἐμ-. This means-that assimilation remains decidedly more 
prominent in the semi-literary MSS of NT than it is in the non-literary 
papyri even of the early Ptolemaic age, but the difference between labials 
and gutturals is still well marked. According to WH (App.? 156 f.), non- 
assimilation is the usual practice in the best NT uncials before x, w, β, 
b; κ, γιχ; 6 73 A, p. But some words have assimilation regularly, the 
ἐν compounds showing it more freely than those of σύν. Details may be 
sought in their list, or more fully in that of Gregory, pp. 73 ff., where 
uncial evidence in each case is supplied. How far the oldest uncials in 
this matter represent the autographs must be left an open question. 
Note the frequency of ἐμ μέσῳ “in good MSS wherever ἐν μέσῳ occurs, 
but never in 8, B, D or D,”; also ἐγ Kava Jn 211 AF, ἐγ γαστρί Lk 2123 A, 
ἐμ πραύτητι Jas 151 &, ἐμ πολέμῳ Heb 115: pl8, σὺμ Μαριάμ Lk 2° AE αἱ. 
σὺμ πᾶσιν Lk 3451: EG al. (Debrunner 14). Thackeray 131 gives corre- 
sponding evidence from LXX, but remarks that such forms were probably 
more abundant in the autographs—which for the NT is hardly likely. 
The rationale of this tendency to drop assimilation 
may be sought partly in lento pronunciation : a nasal 
ending a syllable would naturally tend towards ν, 
the form of all final nasals in pausa. This is supported by the frequency 
of such spellings as ἄνγελος and πένπω (cf. Mayser 235 f., Nachmanson 
106): NT exx. are Kevypeat Ac. 1818 RADE, πένψω 13 times in D,, 
ἔλανψεν 2 Co 4° D, (Debrunner 14), where the ν was not due to ety- 
mological association as in other NT words, παλινγενεσία, πανπληθεί, 
ἐξηρανμένην (Mk 1130 W), μεμιανμένοις (Tit 115 D,). We must note that 
nasals at the end of syllables were by this time becoming faint, and the 
distinction between their three classes (μ, v, y) was less easily heard. It 
should be added that in literary papyri of the period assimilation was 
decidedly more abundant : Mayser (p. 232) gives 12: 8 before labials and 
10 :2 before gutturals in favour of assimilation for a series of classical 
papyri from the Ptolemaic age. The contrast emphasises the non-literary 


practice the more. 
: The form Βεεζεβούλ, which we must assume to 

Βεεζεβούλ be for Βεελζεβούλ (read by all the later Greek 
authorities), is found throughout in 8B except in Mk 3532, where δὶ 
joins the mass. No explanation of this reading is suggested which 
would justify its originality : we must perhaps assume a kind of assimi- 
lation based on the abnormality of the combination A¢ in Greek. Τ so, 
the corruption must go to swell the list of small errors which δὲ and B 
share, proving their common origin. The word involves a further 
problem in the matter of A, if we follow the Latin and Syriac vss. in the 
equation with OT Ba‘al 2‘bab. But this is unnecessary, and whatever 
interpretation is adopted, we may be quite sure that -ὰ is the original 
and -8 a primitive guess by students acquainted with Hebrew. 
; Τολγοθά has also lost X (by dissimilation ?), when 

Γολγοθά, compared with either Hebrew or Aramaic: see ' Βὲ 
s.v. In Mt 2788 A and syrhr have Modyod a. 


Causes of non- 
assimilation. 


106 A GRAMMAR OF NEW TESTAMENT GREEK. [8 42-43 


In 2 Tim 41:8 φελόνην is read practically without 
variant. Its relation to Lat. paenula is difficult, for 
the latter can be quoted from a much earlier date. In papyri the form 
φαινόλης is commoner, but both are found. If it were not for the Latin, 
we might have regarded gay. as popular etymology (daive). It is 
quite possible that both Latin and Greek go back to some unknown 
foreign word. 


Φαιλόνης. 


: The word μογγίλαλος in Mk 7%? is found in W, in 

ee of LNA al., and the important cursives 28 and 33, with 

ae three of the Ferrar group. Since the word μογγός 

hoarse is attested (very rarely) from v/a.D.—see E. A. Sophocles s.v.—we 

must probably regard this as a real v.l., rather than as a nasalising like 
some words mentioned below. 

The insertion of μ᾽ in λήμψεσθαι, λημφθῆναι etc., is a Hellenistic 
analogy-product belonging to the flexion of that verb (see § 95). The 
(classical) loss of » by dissimilation in compounds of πίμπλημι and 
πίμπρημι With ἐν and σύν is overcome by analogy in Ac 14:7 DEP, 288 
cu’ ; as Lobeck shows (Phryn. 95f.), forms with » appear frequently in 
classical MSS. In the latter place N* reads ἐμπιπρᾶσθαι. The inserted 
ν in θηνσαυρούς Mt 21! D is no doubt due to the Latin: we find the 
loan-word often in Plautus spelt thus (see Lindsay Latin Language 69). 
Per contra, in the Roman name Cléméns the nom. (not in NT) was written 
Κλήμης, though cénsus was transliterated κῆνσος. The n in Latin was 
faintly heard in these combinations. WH write ᾿Αδραμυντηνός as the 
adj. of Hadrumetum in Ac 973, following AB 33 and the Bohairic, with 
some minor cursives. 


Sibilants. 


§ 43. It seems probable that in the Hellenistic 
period ¢ had passed from its earlier double value as= 
zd or dz into the soft z as in English ; that is, the MGr pronunciation was 
reached in the Ptolemaic age. A possible survival of zd in the Κοινή might 
be recognised in “A¢wros: in IMAe i. 406 we find ’Ac¢or[is], with the 
note, “‘Semiticam formam Aésdéd optime reddit.” The name presumably 
acquired this orthography before the sign had changed its value. That 
Ny was transliterated Εσδρας in the same early period proves nothing, 


Z in the Kou. 


for the δ could be intrusive as in ‘IodpajyA—see § 41. For the proof of 
this development of ¢ see Mayser 209 and reff. there. It depends largely 
on the frequency with which ¢ in our documents is substituted for o 
before voiced sounds. In NT we find ζβέννυμι in Mt 1239 258 Ὁ, 1 Th 
519 B¥D*, FG, ἄζβεστος Mk 948 N; ζμύρνα Mt 211 DW, Jn 19% δὲ (c¢p.) 
DsuppW ; Ζμύρνα Rev 11! 28 δὲ and Latin. Soin BGU iv. 1175 ter (B.c. 4) 
Ζμύρνα is the name of a woman of Persian birth ; in P Oxy viii. 108889 
(early i/A.D.) (uvpva occurs in a medical prescription ; and in P Ry] ii. 
15318 (mid. ii/A.D.) we read ἐν Ζμύρνῃ τῆς ᾿Ασίας. We find Zmyrna in 
Catullus 95, and elsewhere in MSS of Latin authors. Lightfoot’s note 
(ii. 331) on Ep. Zgn. ad Polyc. (ad init.) gives abundant evidence that in 


§ 43] SOUNDS AND WRITING. 107 


the name of Smyrna the ¢ and the o were used impartially in ii/a.p. 
(init.). ‘In the earliest coins the ¢ seems to be preferred, in the latest 
the o.” Ifso, ¢ would rule in 1.4.0. Crénert Mem. 95 gives a multitude 
of parallels from the Herculaneum papyri (before a.p. 79) and from 
inscriptions. It is in fact abundantly clear that if ἐμ and ¢8 are 
“Western” only (WH App.? 155), that is not so much a reason for 
banning the ¢ as for approving the “ὃ Text” in its (characteristic !) use 
of spelling that was coeval with the autographs and true to the pro- 
nunciation. In ΜΟΥ of and op are pronounced with voiceless ¢. 

The relations of oo and rr may be treated here: 
it must be remembered that in earlier Greek oo(rr) 
nearly always, and ¢ very often, arise from a guttural followed by con- 
sonantal y. In MGr we find in Carpathos and Chios oo (and cy) 
becoming ro, while “the transition from ¢ to d¢ (e.g. παίάξω -- παίζων is 
more widely spread” (Thumb Handb. 22). One is tempted to regard 
both as survivals, and take ts as the old pronunciation of oo (perhaps ¢th 
of rr), just as we know dz was that of ¢in class. Greek. Without dwell- 
ing on this conjecture, we go on to note that the Kowy has oo almost 
exclusively where Attic had tr, which was hardly used outside the 
contiguous districts of Attica and Boeotia. The following instances of rr 
are accepted by WH in NT: κρεῖττον in Paul 1/4, 1 Pet 1/1, 2 Pet 1/12, 
κρείττων etc. Heb 11/12 and one doubtful; ἐλάττων 2/4, ἐλαττόω 3/3, 
ἐλαττονέω 1/1; ἥττημα Paul 2/2, ἡττάω 2 Pet 2/2, but ἧσσον Paul 2/2 and 
ἡσσώθητε Paul 1/1; ἐκπλήττω 1/12 (in Ac 131%). These are mostly 
explained satisfactorily by Wackernagel’s thesis referred to above (§ 42). 
The verb from ἥσσων was either ἑσσόω (Ionic) or ἡττάω (Attic): when 
the Κοινή took it over, it naturally did not fuse these alternatives into 
ἡσσάω. Ἥττημα accordingly was an Attic word, adopted as it stood. 
Wackernagel thinks that ἥττων followed ἡττᾶσθαι, and in its turn in- 
fluenced its synonym ἐλάττων and its antithesis κρείττων, both of which 
appear in ΤΧΧ-- ἐλάττων with its derivative verbs greatly outnumbering 
the oo form. Here the process was helped by the fact that the verb 
ἐλαττόω (and noun -wors) was specifically Attic and had no Ionic rival to 
endanger its rr. No special explanation can be suggested for the isolated 
ἐκπληττόμενος, With which ef. καταπλήττεσθαι in P Petr ii. 46}. 18. (¢, B.o, 
246) ;1 but similar isolated or occasional variations can be quoted from 
papyri and other Kown documents—see Wackernagel’s samples. The 
influence of Attic was so commanding that we cannot be surprised if 
reminiscences of an Attic peculiarity slip in to disturb normal pro- 
nunciation. 


Xo and rr. 


There remains the question of transliteration from 
Semitic, already raised on the name”A¢wros. Burkitt 
shows? that Na¢apa (Ναζαρέτ) stands practically alone 
if we make its ¢ represent a Semitic y; for which reason among others 
he seeks to get the adjectives Ναζαρηνός and Ναζωραῖος from 31. Normally 


{ in trans- 
literation. 


1 Wrongly given from P Grenf in Mayser 223. 
3 Syriac Forms of NT Proper Names (Brit. Acad. 1912), 16, 28f. 


108 A GRAMMAR OF NEW TESTAMENT GREEK. [8 43-44 


¥ was o and ἢ was ¢,as we might expect, δ᾽ and wy being also a, since 
Greek had no means of differentiating. 


Gutturals. 


§ 44. The combinations yp and yy were pronounced 
ΠΑΡ. with y as ng (as in kingmaker ete.). In lonic during 
iv/B.C. τιγν- became -iv- in γίνομαι and γινώσκω, perhaps through dis- 
similating influence of the initial y (so Brugmann-Thumb Gr, 126): 
Thumb thinks that the same forms in Doric, Thessalian and Boeotian 
arose from later Attic (B.c. 300 and after) and Hellenistic influence. 
These forms are universal in the Κοινή, as innumerable papyrus records 
show—a few pedantic revivals of yyy. serve as exceptions to prove the 
rule. It is curious that W, which has so many thoroughly vernacular 
spellings, uses γιγνώσκω often, and γίγνομαι, though less frequently 
(Sanders 23). Sporadic instances of the dropping of ng may be seen in 
Adpvé Rom 313 AP 33 cu’, and σάλπιξ 1 Co 148 ALP. The papyri show 
innumerable instances of ἐκ with « softened before 
voiced sounds: thus the recurrent formula καθάπερ 
ἐγ δίκης. Mayser (225 f.) shows that it is normal, « being quite exceptional. 
In NT it is curiously absent, but note ἔγγονα 1 Tim 5* D* cui (see reff. in 
Mayser 228), ἀπεγδύσει Col 21! B*, ἀνέγλιπτος Lk 1258 D, ἐγλύου Heb 12° 
p'8 (from Debrunner 14), also ἐπιδιγνύμεναι in Ac 9°° A. So yvadeis, 
which in old Attic was κυ: yv. is found in an Attic inser. of iv/B.c., and 
exclusively in Ptolemaic papyri (Mayser 169f.). It may be noted here 
that ἐκ sometimes combined with initial o to €: thus 
eZoy=ex σοῦ Mt 2° NC al., 2119 Dal. Soin papyri 
ἐξυμφώνου etc. (VCR xv. 31). The v./. in Ac 2759 
ἐξῶσαι NAw, ἐκσῶσαι BC boharm, is really perhaps a mere matter of 
orthography in its inception. 

On ὄρνιξ see § 55 (3) (2). Πανδοχεύς -eiov has x 
from the influence of its original δέχομαι : the κ (ef. 
Tonic δέκομαι) is still found in Syll. 901 (? i/B.c.) and even in P Gen-54?6 
(iii/a.D.). For « in Lk 10% stand X*z, 0.85 N*D*, against ABLWw : 
Ti takes the correction in ND as evidence that the «x form was obsolete. 
Forms like ex@pds—common in D1—and the converse ἐχθέσει (Wis 1114, 
from ἔκθεσις : see Thackeray 103) show the development of @ into a 
spirant, which involved a strange articulation of the guttural before it. 
Metathesis of aspiration produces interchange of y and « in χιτών and 
other words : see § 40. 


K and y. 


Contiguous 
κ and o-. 


K and x. 


Semitic words show yx occasionally as a trans- 
literation of the gutturals, including even &. So 
“Ακελδαμάχ (ax- BCD, ay- 8A) = Aram. xq Son, 
and Sepax=NVWD- See Kautzsch Gramm. d. bibl. Aram. (1884) 8. 


Dalman (Gramm. 161) compares Ἰωσήχ Lk 3?°=spy, and says that 


Semitic words 
with x and x. 


1 See Rendel Harris, Codex Bezae 141. Also οἵ. Mayser 172. 


§ 44-45] SOUNDS AND WRITING. 109 


the x marks the word as an indeclinable. Elsewhere δὲ always dis- 
appears H aBBa=N3X, σίκερα --λ 5} ete. But ἐφφαθά-- ΠΗ͂ΞΙΝ (Kautzsch 
10), while Π and ΠΟ initial are often only a (presumed) rough breathing. 
“Ῥαχάβ in Mt 15 stands for 3m, which in Jas 2”, Heb 1151 is Ῥαάβ, all 
without variant. In Ῥοβοάμ for Dyan both y and n disappear. The 
transliteration of y by y under certain conditions is rather a matter for 
Semitists than for us: it is enough to observe that in our period y was 
generally a spirant, so that Touoppa represented my satisfactorily. The 


instances of x standing for p (cf. ἀχελδαμάχ above) may be due to the 
differentiation of p and 3: χ when spirantised may have represented p 
better as being articulated farther back. An instance of the opposite 
procedure is where Kavavaios represents an initial p, and Xavavaia a 3, on 
the evidence of the Syriac.! The representation of m by « in Κλωπᾶς 
cannot be accepted if it means direct transliteration, since Ἔδπ has 


«Αλφαῖος already to represent it ; but Dalman (Gramm. 142 n.8) suggests 
that it might be the Gentile name of a Jew spbm—as a Saul took the 
similar-sounding “ Paul,” or a Silas “Silvanus.” This presumes our 
identifying Κλωπᾶς with Κλεόπας, on which see above, ὃ 38. 


Labials. 


§ 45. Σφυρίς for σπυρίς is well attested in NT, and 
appears very often in papyri: CR xv. 33, xviii. 107, 
Mayser 173. So oddyyos Mk 15°° 1), σφεκουλάτορα Mk 657 W. See 
Crénert Mem. 85n.°. Meisterhans® 78 shows that σφόνδυλος was Attic in 
iv/s.c. Healso cites Attic inserr. of ii/a.D. and later for ᾿Αφφιανός, where 
φφ is assumed to represent Latin pp, as in Phm? ’Addia D*. So from 
mid. ii/B.c. the Latin Sulpicvus becomes Σολπίκιος. (Contrast ᾿Αππιανός 
novies in P Oxy i. 33 (ii/a.D.): so CP Herm 127 vs 111.5 (? ii/a.p.), and 
BGU iii. 785! (i/a.D.).) But the inscriptional parallels from Asia Minor 
in Kretschmer’s Hinleitung 346 f. make it highly probable that this name 
in its various forms was affected by an Anatolian word (cf dma “ papa” 
in Greek nursery speech) unconnected with Appius. This last appears 
in Ac 28, where 33 cu? read’Arqdiov. No doubt in Asia Minor this 
word influenced the spelling of really Latin names. Schweizer Perg. 
110 remarks that when ¢ became a spirant it probably remained bilabial, 
and therefore not quite equivalent to the f of imperial Latin, which like 
ours was labiodental. The relations between π᾿ and ¢ are concerned in the 
v.l. at Ac 1318 ἐτροποφύόρησεν NBDw, erpod. AC*E 33 latvt syr vg hl sah 
boh ; there is no probability that the former could arise from the latter 
phonetically. See § 28. 

The addition of ¢ in’Acad Mt 17 NBC p! (also D in Lk) 1 ete. latvt sah 
boh is discussed by F. C. Burkitt in Proc. Camb. Philolog. Soc. for March 
4, 1897. He shows from Latin evidence that the true LXX form of the 


for π. 


18, Ο. Burkitt, Syriac Forms of NT Proper Names (Brit. Acad. 1912), 5. 


110 A GRAMMAR OF NEW TESTAMENT GREEK. [ὃ 45-46 


name, however explained, was ᾿Ασάφ, so that Mt was following the Greek 
Bible: this disposes of Salmon’s criticism on the NT uncials (Some 
Thoughts on the Textual Criticism of NT 29, 156). 


An important point for Greek pronunciation is 
B and o(u) for a 


ais mI } De - 
w sound. raised by the transliteration of 1) and Sclvanus. Un 


cials have only Aav(e)id (often abbreviated), and Δαβίδ 
comes in with the cursives. The earlier spelling is exactly parallel with 
the representation of Roman names like Flavius, for which in the papyri 
we find Φλαύιος as well as Φλαούιος : thus P Oxy 11. 237%"! (a.p. 186), 
vi. p. 223 (and 356) (A.D. 154), 991 (A.D. 341)—the spelling with ov is 
much thecommoner. The case of Silvanus is rather different, as v follows 
a consonant. Σιλουανός is the overwhelmingly attested form, but Σιλβανός 
has better warrant than Δαβίδ, appearing in B at 1 Pet δ13 and in DEFG 
at 2 Co 119,1 Th 11, 2 Th 1! (where also add 424**). Now in the 
papyri Σιλβανός stands alone from iii/a.D. onwards: the very few earlier 
instances of the name quotable from inscrr. and papyri seem to show that 
the 8 spelling did not arise (with one exception, which is not quite certain) 
till the end of ii/a.p.1 That δὲ always and B 3/4 times should have pre- 
served the long obsolete ov of the autographs is another of the striking 
trifles which go to prove their accuracy. It was only in later times that 
8 became the normal representative of F (i.e. Eng. w): being a spirant 
like our v (but interlabial), it did not accurately express the Latin wu or 
Hebrew }. Earlier Egyptian contacts of 8 and v may be seen in Mayser 
115, where from ii/B.c. words like ῥαύδους = ῥάβδους (P Par 4058), ἐμβλεύ- 
σαντες (P Lond 23!5=i. p. 38)=-8dép-cavres made up afresh with d=f, 
etc., show the spirant value of both. Note that o sometimes represents 
Latin consonant ιν, as Koivros, ᾿ΟΘαλέριος. 


Dentals. 


9 § 46. Interchanges of 6 and 7 from metathesis of 
T and ὑ. aspiration are treated in §40. Μαστός varies: Lk 1157 
μασθοί DG 2329 D*FG (μαζοί C), Rev 118 μασθοῖς N (μαζοῖς A). Maos 
and μασθός are primitive doublets : cf. Skt. meda- and médha- (Brugmann- 
Thumb Gr. 117, 125). Μαστός (the normal cl. form) may perhaps be 
independent : it might be a verbal from the root (cf. padaw) from which 
the other two are derived. See the material in Walde, Lat. Etym. 
Worterb2 453. Φόβηθρον Lk 21)! BDW (Is 1917 B) is presumably due 
to the analogy of words like κόρηθρον, κύκηθρον (W. F. Moulton in WM 
119 n.2): add κάλλυνθρον BGU iv. 1120" (B.c. 5). See Brugmann- 
Thumb Gr. § 206, and Thackeray 104 (uac@ds and φόβηθρον in LXX) 
Mayser 179 illustrates the frequent change of σθ to or in papyri. Rendel 
Harris describes this as “very common in Codex Bezae”: he instances 
Mk 41 καθῆσται, Ac 1955 ἐπίστασται. 


1 See the facts concerning Silvanus in Vocabulary, s.v. It must be noted 
that in some names β represents Latin wu much earlier: see Viereck Sermo 57, 
with instances from i/z.c. Livia is AcBla in P Byl ii. 1276 (a.p. 29). 


§ 46] SOUNDS AND WRITING. 111 


Gabeie ate The variation between οὐδείς μηδείς and οὐθείς 

>" unOeis is one of peculiar importance as a test of our 
MSS : οὗ what is said on ἐάν for ἄν in ὃ 38. The history of this rather 
shortlived development is acutely traced by Thackeray (pp. 58-62 and 
104f.), whose account should be carefully followed. That οὐτεμία is 
never found proves that mixture of οὔτε and οὐδέ has nothing to do with 
it: there has been a re-formation οὐδ᾽ eis, with 6+h producing 6—see 
Brugmann-Thumb 170f. “First found in an inscr. of 378 B.c.,! it is 
practically the only form in use throughout the Greek-speaking world 
during iii/B.c. and the first half of 1i/B.c. In 132 B.c. the ὃ forms begin 
again to reassert themselves, and the period from that date to about 100 
B.C. appears to have been one of transition, when the ὃ and 6 forms are 
found side by side in the same documents. For i/B.c. we are in the 
dark, but in i/a.D. we find that οὐδείς has completely regained its ascen- 
dancy, and by the end of ii/a.D. οὐθείς, which still lingers on in i/ii a.p., 
mainly in a single phrase μηθὲν ἧσσον, is extinct, never apparently to 
reappear, at all events not within the period covered by the papyri.”? 
It follows naturally that οὐθείς plays a very small part in NT. It is 
accepted by WH eight times: Lk 2255 2314, Ac 159 1927 2676, 1 Co 132, 
2 Co 118 (ov6.), and Ac 2758 (un8.). To these might be added Ac 2033 
NAE 27%4 A, 1 Co 133 NA 33 cu? : since -θείς was obsolete long before our 
oldest MSS, we should incline towards accepting it as often as good 
uncials show it. But against these possible 11 instances we have some 
220 of οὐδείς and 80 of μηδείς. We have to note the peculiar case of the 
verb ἐξουθενεῖν, a verb coined while οὐθείς was still in use : Thackeray 
shows (104 f.) that ἐξουδενοῦν was coined afresh when οὐδείς was reassert- 
ing itself, and mixed forms -δενεῖν and -θενοῦν appear in a few places in 
LXX. In Plutarch we have ἐξουδενίζω, perhaps a corrected form from 
ἐξουθενίζω which we have ina scholiast : ἐξ]ουδενῆ[σαι is said by Schubart 
to be the most probable reading in BGU iv. 11173! (B.c. 13), and it is 
printed by Mitteis in Chrest. 11. p. 129 without question. In NT ἐξου- 
θενεῖν, as a long-established word which has detached itself from its 
origin, stands without doubt in 11 places. In Mk 912 there is doubt : 
ἐξουθενωθῇ & 69 is confronted by ἐξουδενηθῇ in BD 565, while LNW have 
ἐξουθενηθῇ and ACXA al. eovdevw6;—perhaps the most probable reading, 
since the general NT form -θενη- will explain the 6 of the δὲ and the η of 
the BD reading. From other places the only variants are Lk 23" -aoas X, 


1 But in Amer. Journ. of Archaeol. vii. 152, S. O. Dickerman gives an inser. 
from Cleonae which has μηθέν, and he assigns it to early v/B.c. at latest. I 
cannot criticise this judgement. 

* Thackeray Gr. 58. Since this was written, our ‘‘darkness” as to i/B.c. 
has been somewhat relieved by the publication (in BGU iv.) of about a hundred 
papyri from Alexandria, dated under Augustus. One of these documents, No. 
1141, a private letter, and almost the only paper in the collection which is 
not formal, shows οὐθέν and μηθέν once each, and even this has μηδέν twice. 
Even the formal character of the remainder does not altogether discount the 
fact that οὐδείς appears 5 times and μηδείς 56 (largely in identic formulae). 


112 A GRAMMAR OF NEW TESTAMENT GREEK. [S 46-47 


-isas W (cf. Plutarch’s verb, unless it is mere itacism—see § 34), Ac 4!! 
-wbeis οι (-δενωθείς cu'), 1 Co 178 ἐξουδενωμένα 33 (-θενω- cul, 1611 
ἐξουθενώσῃ 38 (-δενω- cu’), 2 Co 1010 ἐξουδενημένος B (as in Mk 9”). 

A few miscellaneous variant forms may be noted. 
A dental is dropped in ἄρκου Rev 13? (all une.). It 
is explained in Brugmann-Thumb 151 as an effect of 
popular etymology, produced by dpxéw and τὸ ἄρκος “defence.” The 
link hardly seems obvious: if we are seeking an etymon it is more 
plausible to try ἄρκυς, the bear being assimilated in name to the net that 
snared him. But Boisacq (s.v.) regards both forms as primitive Idg 
doublets. Add the old word dpxados “ young panther.” ”Apxos is both | 
literary (Aelian) and vernacular Kown (see Vocab. s.v.), and MGr. For 
the added dental in σφυδρά (Ac 37 N*AB*C*) we have only Hesychius 
to quote. (Ανα)βαθμός came into the Κοινή from Ionic, according to 
Phrynichus ; but G. Meyer (p. 365) gives inscr. evidence against this (cf 
Thumb -Hellen. 73). Anyhow it is Hellenistic, while Bacpos is Attic. 
See Vocab. on both words. On the other hand no trace survives of the 
old form ὀδμή, which figures in Herculaneum papyri and some later 
writers as v.l.: see Crénert Mem. 136, who notes dissidence among the 
old grammarians. ᾿Εθύθη 1 Co 57 is only a blunder in very late sources 
of TR. Aandrare concerned in the Hellenistic ποταπός, which comes 
from class. ποδαπός (see under Word-formation) by assimilation to πότε. 
There is no connexion with the frequent substitution of + for δ in badly 
written papyri, which is due wholly to Egyptian native pronunciation. 

A curious substitution of x for @ occurs in D at Mk 651 γενεχλίοις. 
There is only a fortuitous resemblance to the χ in ὄρνιξ. 

In foreign words there is some wavering between 
τ and 6, shown by doubling (as Ma6@aios) to be 
spirant, representing Aram. ἢ. So esp. Ναζαρέτ and 
Ναζαρέθ, the latter predominating in Mt and Lk, the former in Mk and 
Jn, according to Gregory 120. WH print only τ (Na¢apd Mt 41%): the 
form with 6 is attested by δὲ 4 times, B at least 4, D also 4. W oddly 
drops it in the best-attested place, Mt 2111, and in Lk 4 times, but has it 
everywhere else, as the later MSS normally. 


Miscellaneous 
variations. 


In foreign 
words. 


Movable Letters 


᾿ 8 47. (1) Final -s in οὕτως is practically fixed. Οὕτω 
Final s movable. ;. eee WH 10 times (wv. Ἰὰς πο in WM 
44n.) out of over 200, on the mechanical principle of accepting an omission 
found in δὲ or B supported by A or C—a principle they adopt also for 
movable -ν. In Ptolemaic papyri οὕτω is found a few times, even before 
vowels, but οὕτως predominates (Mayser 242f.). On the other hand, 
Crénert counts 75 instances of οὕτω to 25 of οὕτως before consonants, 
in the Herculaneum rolls included in his survey (Mem. 142). Since 


1 Their neglect of D in this matter is in accord with their general principle, 
but it is hard to defend it to-day. 


§ 47] SOUNDS AND WRITING. 113. 


these are literary, they need not disturb the impression that οὕτως is 
normal. 

”Aypcand μέχρι were Attic, according te Thomas Mag. (p. 135) and 
Phryrichus (al.—see Rutherford NP 64): Moeris (p. 34) calls ἄχρις 
Hellenistic. But the record of the forms with -s is not good enough to: 
justify this claim. They appear first in the Roman period (Mayser 243) : 
an early instance is BGU iii. 830!° (i/a.D.) ἄχρις ἄν. Instances betore: 
vowels appear in early Fathers (Reinhold 37). In NT “ ἄχρι usually 
precedes vowels (14-16 times), Gal 3!° ἄχρις ἄν or οὗ being the only 
certain exception: μέχρι preceding a vowel is certain only Lk 16!°, 
μέχρις 2-3 times” (WH App.? 155). They give ἄχρις as alternative in 
Rom 1155, and read ἄχρις in Heb 518, μέχρις in Heb 1325 with alternatives, 
in Mk 1399 Gal 419 without alternative. 

λντικρυς Χίου is found in “ all good MSS” at Ac 3015 (WH). 

The omission of -s in -κις adverbs (Cronert Mem. 142 f.), not 
uncommon in the Κοινή, has no place in NT. 

Radermacher (Gr. 39) observes that final -s and final -ν were alike 
feeble in the Hellenistic period. ‘There are even instances of -c written 
for -s,as more often for -ν : thus τὰς dda P Tebt i. 61 ὁ 37% (Bc. 118). 
See Mayser 136. 

(2) Final -ν (ν ἐφελκυστικόν) is so universal in 
the forms which admit it at all, that it is only 
necessary to take note of omissions. Modern use, by which ν is inserted 
before vowels only, is known to be wrong even for classical writers, and 
in Hellenistic it is altogether to be set aside. Indeed a superfluous -ν 
appears largely in forms which had never known it. A conspicuous 
instance is ἦν, subj. of εἰμί, which has misled even Deissmann (LAF 155). 
See Prol. 49, where μείζων acc. sing. is quoted from Jn 5°° AB al. (add 
W). Cf. βορρᾶν gen. in δὲ in LXX (Thackeray 143). Cf. from papyri 
P Oxy iii. 505 (i1/A.D.) ἀπηλιώτουν, Preisigke Samim. 431714 (c. A.D. 200) 
ev ᾿Αλεξανδρίᾳν, P Tebt i. 104 (i/B.c.) ᾿Απολλωνίᾳν, P Oxy vill. 10884* 
(i/A.D.) εἶταν, P Ryl ii. 90%? (iii/A.D.) τῷν ἡμῶν κινδύνῳν, tb. 160% 
(A.D. 28-9) β]εβαιώσιν (dat.) etc. ete. Cf. Nachmanson Beitrége 66 f. 

For the practice of NABC in the matter of movable -v, see WH App.? 
153-5, who explain there the admittedly mechanical rule by which they 
decide whether to print -» or omit it ; see under (1) above. 

The irrational addition of -ν may be set beside its irrational omission, 
for which see many exx. in Mayser 190f. One recurrent instance may 
be named, πάλι for πάλιν, a vulgar by-form found in post-Ptolemaic 
nserr. and papyri—ef. Mayser 241. It occurs in W at Jn 1%, 

Final -ν has the same uncertain tenure in MGr that it had in 
Hellenistic, and the range of its variation has been considerably 
extended. See Thumb’s account, Handbook 24 f. 

(3) Final -c after ἡ or o—the c subscript of mediaeval 
and modern writing (see above, ὃ 27)—was in 1/B.c. 
and i/A.D. inserted and omitted so freely in papyri that it may be 
counted as a movable final indifferent for pronunciation, even beyond -ν. 
A finely concentrated instance is BGU iii. 883° (ii/A.D.—by which time 

VOL. 11.—8 


Final -v. 


Final . movable. 


114 A GRAMMAR OF NEW TESTAMENT GREEK _[S 47 


it was rare) ὀκτὼι, οὐλὴι δακτύλῳ πρώτωι : the « is in these four words 
inserted once rightly, twice wrongly, and once wrongly omitted. It 
might be substituted for the other weak finals, as we have seen : add 
such exx. as BGU iv. 1188'* (B.c. 15) εἰς τὴν κώμηι, and conversely 
Preisigke Ostr. 157 (B.c. 59) ἀριθμῷν dat. sg. Further instances are given 
in CR xviii. 108a. See in general on the long diphthongs in § 36. 
Since -a@ might be read -d and so pronounced -e, the irrational -: was 
naturally added to -a less often than to -y or τῳ. The insertion of this 
irrational « is best taken as a mere consequence of literary tradition: the 
proper limits of a now functionless letter were forgotten, and indiscrimi- 
nate insertion paved the way for impartial omission from ii/a.D. Accord- 
ing to Gregory Prol. 109 the c is found very rarely in NT uncials : he 
quotes ἤιδισαν (-ΞΞ ἤδεισαν) Mk 1°* D, ὧι Mt 2515 U, Lk 7* A, and ξύλωι 
Lk 2331 K. He adds on Scrivener’s authority that it is not found sub- 
script in the minuscules before the time of cod. 71 (written a.p. 1160). 
(See §§ 27, 36.) 


A GRAMMAR OF 
NEW TESTAMENT GREEK 


Dr. J. H. Moulton’s 


Grammar of New Testament Greek 


“marks a revolution in New Testament study.” 
Dr. J. Hastines in The Expository Times. 


VOL. I. THE PROLEGOMENA. 
CONTENTS :—General Characteristics—History of the 
‘““«Common” Greek—Notes on the Accidence— 
Syntax: The Noun—Adfjectives, Pronouns, Pre- 
positions—The Verb: Times and Modes of Action 
—Voice—Moods—The Infinitive and Participle— 
’ Additional Notes—Indices to Quotations, Greek 

Words and Forms, Subjects. 


VOL. II. ACCIDENCE AND WORD-FORMATION. 
Part i. General Introduction. 
Sounds and Writing. 
Part ii. Accidence. 
(a) Nouns ; (ὁ) Verbs and Lists. 
Part iii. Word-Formation, with Appendix on Semitisms, 
and Indexes to Vol. IT. [lz Preparation. 
When completed, the Three Parts of Vol. 11. 
may be bound together in cloth binding. 


VOL. III. SYNTAX. 


Based on Dr. Moulton’s own researches, and following 
his method of first-hand investigation of the Papyri, and 
other sources of our knowledge of the Koine. 

[71 Preparation. 


Current Prices as in Messrs. Clark's latest Catalogue. 


EpinpurcH: T. ἃ T. CLARK, 38 GEORGE STREET. 


A GRAMMAR OF 
NEW TESTAMENT GREEK 


BY 


JAMES HOPE MOULTON 


M.A.(CanTas.), D.Lit.(Lonp.), D.D.(Ep1n.), D.C.L.(DUNELM.) 
D.THEOL.(BERLIN AND GRONINGEN) 


LATE FELLOW OF KING’S COLLEGE, CAMBRIDGE 
LATE GREENWOOD PROFESSOR OF HELLENISTIC GREEK AND INDO-EUROPEAN PHILOLOGY 
MANCHESTER UNIVERSITY 
LATE TUTOR IN NEW TESTAMENT LANGUAGE AND LITERATURE 
WESLEYAN COLLEGE, DIDSBURY 


Vol. II 
ACCIDENCE AND WORD-FORMATION 


Part ii 
ACCIDENCE 


EDITED BY 


WILBERT FRANCIS HOWARD 
M.A.(Mancu.), B.D.(Lonp.) 
TUTOR IN NEW TESTAMENT LANGUAGE AND LITERATURE 
HANDSWORTH COLLEGE, BIRMINGHAM 


ΕΡΙΝΒΌΒΟΗ : T. & T. CLARK, 38 Gezorce STREET 


[920 


PRINTED BY 
MORRISON AND GIBB LIMITED, 
FOR 
T. & T. CLARK, EDINBURGH. 


LONDON ; SIMPKIN, MARSHALL, HAMILTON, KENT, AND CO. LIMITED 
NEW YORK: CHARLES SCRIBNER’S SONS. 


PREFACE TO PART ii. 


=< 


THE editor’s task in preparing the second part of this volume 
for the press has been comparatively simple. The manuscript 
was completed by Dr. Moulton, except that spaces were left 
for the numerous cross-references to be inserted later. The 
editor is also responsible for the verification of all other refer- 
ences and for a few alterations in statistical data. One other 
interference with the manuscript should be confessed. A 
noticeable feature of this grammar is the use of bold letters 
in the noun and verb paradigms to distinguish forms or types 
which actually occur in the New Testament While the 
sheets were passing through the press it became evident that 
Dr. Moulton could not have revised this part of the work. 
The editor has spared no pains to make that distinctive record 
as accurate as all else that came from the author’s pen, and 
he makes grateful acknowledgment of the patience with which 
the printers have reset so many of the paradigms. The Rev. 
Dr. J. G. Tasker has kindly read through most of the proofs ; 
and no thanks are warm enough for the unfailing generosity 
of Professor Milligan, who has not only kept a watchful eye 
on every page at each stage of proof revision, but has himself 
tested every papyrus reference, in addition to making many 
fruitful suggestions. 

A double sadness weighs upon the writer as he calls to 
mind the promise of assistance for Part iii. which he 
announced in the Preface to Part i. An appendix on Semi- 
tisms in the New Testament will appear, but it will not be 
written by that brilliant scholar who had promised Dr. 


ν 


vi PREFACE. 


Moulton this contribution to the Grammar. Mr. Bedale died 
in a military hospital at Cambridge before the publication of 
Part i. Mr. Henry Scott lived just long enough to see it in 
print. The death of this modest and laborious scholar is a 
grievous loss to all who had reason to value his incomparable 
accuracy in all the minute details of statistical research within 
the field of Biblical Greek. During his last illness his chief 
concern was for the success of this work, and he bequeathed 
his manuscripts and note-books to the editor in the hope that 
Dr. Moulton’s great project might the better be carried to 
completion. 

This sacred task becomes the heavier, with first brother- 
in-law and then kindly old friend called from one’s side so 
early in the day. But other helpers will share in this service, 
and there is good hope that before long Volume II. will be 
concluded, and that in due course the volume on Syntax, 
begun by Dr. Moulton, will see the light. 


W. F. HOWARD. 


HANDSWORTH COLLEGE, 
September 1920. 


CONTENTS. 


DEFINITE ARTICLE ., : : ὅ 


Nouns— 
First Declension , 
Second Declension 
Metaplasmus and ἘΠ 
Third Declension ; 
Declension of Semitic Names 
Declension of Non-Semitic Names . 
Miscellaneous Foreign Words and eo ΠΕΣ: 


Latin Nouns ᾿ : 
ADJECTIVES . : : é F 
ADVERBS 


Comparison of Adjectives and Ravarbe 
NUMERALS 
PRONOUNS 


VERBS . : : 
Conjugation and Tense Stems 
Augment and Reduplication : 
A. Present Stems— 
(a) Thematic— 
Active Voice . : ὴ z 
Middle and Passive Voices . : 
(0) Unthematic— 
Active Voice . 
Middle and Passive Voices 
B. Strong Aorist Stem— 
Active Voice . : a 
Middle Voice . 3 
Passive Voice. 


PAGE 


117 


117 
120 
124 
128 
143 
150 
152 
155 
155 


163 
164 


167 
178 


182 
183 
188 


193 
197 


201 
205 


208 
212 
213 


vili CONTENTS. 


C. Weak Aorist Stem— PAGE 
Active Voice . F ‘ : : 915 
Middle Voice. : ‘ ‘ : 4 AE 
Passive Voice . : : : : BG 

D. Future Stem— 
Active and Middle Voices. : ; e218 
Passive Voice . : i : : 219 

HE. Perfect Stem— 
Active Voice . : : 2 , ~ 220 
Middle and Passive Voices . ; Z 5 PB 
F. Verbals : ; : : ξ 5 . 224 
List OF VERBS : - : ᾿ A ᾿ 294 

PARE 4. 
CORRIGENDA. 


P. 16, 1. 14, read Aramaic. 

P. 19, n.2 1. 3 ab imo, read Lietzmann. 

P. 44, § 25, inset title. Under Consonants, insert Mutes, 
P. 91, 1. 10. For § 53 read § 52. 

P. 99, π.1 1. 2 ab imo, first word, read that. 

P. 100, 1. 6. For ὃ 76 read § 77. 


PART II. 


ACCIDENCE. 


Tue Noun and Verb paradigms that follow are printed in bold type 
(as ἡμέρα) whenever the forms themselves or forms on the same 
model actually occur in NT. Small type is used, as κύνα, when the 
word is isolated and only part of its flexion occurs, or when from 
actual instances in NT it is not possible to make up the flexion of a 
model word. When there is reason to doubt the continued existence 
of an inflexional type in Hellenistic, a blank is left. 


VOL. II. PART Il.—9 


PARE SIL 
ACCIDENCE. 


DEFINITE ARTICLE. - 


§ 48. The forms of the Article should strictly be pre- 
sented among the other Pronouns, to which it belongs by 
historical syntax. But in post-Homeric Greek, except for a 
few uses which will be treated under the Syntax, the Article 
had detached itself for special functions answering generally 
to those of our own the; and convenience demands that it 
should be given here. There are no irregularities. The base 
is to-, ta-, except in N. sing. mf. where it is so, sd (as in 
Skt, Germanic etc.): this has in Attic and other dialects 
infected the plur., which was originally τοί. 


Sing. N. 6 ἡ τό Plur. οἱ αἱ τά 

A, τόν τήν τό . Tous τῆς Ta 

G. τοῦ τῆς τοῦ τῶν τῶν τῶν 

D. τῷ τῇ τῷ τοῖς ταῖς τοῖς 
NOUNS. 


First DECLENSION. 


§ 49. This declension includes nouns and adjectives with 
stems in -d, masculine and feminine, together with those 
(distinct in their origin) which in Greek show a nom. sing. in 
-1a, feminine only. The feminine type is original in the -a- 
nouns also: we shall present it first. 

A, Feminine Nouns in -4, -n, and -d. 


ἡμέρα day. φωνή voice. γλῶσσᾶ tongue. σπεῖρα cohort. 


Sing. NV. ἣ ἡμέρα φωνή γλῶσσα σπεῖρα 
A. τὴν ἡμέρων φωνήν γλῶσσαν σπεῖραν 
Ο. τῆς ἡμέρας φωνῆς γλώσσης σπείρης 
2. τῇ ἡμέρᾳ φωνῇ γλώσσῃ σπείρῃ 

Plur. N. αἱ ἡμέραι | φωναί | πὶ εἰν 
4. τὰς asp φωτός all other First Declension words, 
G. τῶν ἡμερῶν | φωνῶν | Gen, plur. always perispomenon. 
1). ταῖς ἡμέραις }) φωναῖς 


117 


118 A GRAMMAR OF NEW TESTAMENT GREEK. [Ἃ8 49 


The Vocative is identical with the Nominative through- 
out. 


It will be seen that Hellenistic agrees with Attic in its treatment of 
ἃ pure (see above, § 33). There are two divergences from Attic in this 
flexion, the first of which affects nouns in -ns as well. 

(1) In spelling the dat. sing. in » was unchanged, but in reality it was 
new. Attic 7 was pronounced as εἰ (é close as in day), while ἡ was the 
open ὃ in there. Hence nom. and dat. sing. of barytone -y nouns were not 
mere graphic variants, as in Hellenistic, where ἡ and » were alike close 
é, while εἰ was ὃ (as in machine). The new dative came from gen.— 
ἡμέρας : ἡμέρᾳ : : φωνῆς : φωνῇ, aided by the uniformity thus established 
throughout Decl. I. by nom. and acc. having the same relation to the 
dative in -ἃ (-as) and -y (-ns) nouns alike. 

(2) There is a partial levelling of -@ stems: nouns in -pa and parti- 
ciples in -via follow γλῶσσα in gen. and dat. sing. 

(a) The evidence of NT MSS is as follows. Mayaipns (-n) ὃὲ 8, B 4, 
A 8,0 , Ὁ 3, Dy 2, L 3, A 4, W 8, p!* 3: add single occurrences in T 33 
81 124. πλη(μ)μύρης 1 in NB*LWE 33. Πρῴρης + in NA 338. Σπείρης 
3 (Ac) with no serious variant except B in 10! and P twice. Samgeipy + 
NAEP. Συνειδυίης } NABE. In the papyri both -pys and -vins are 
normal A.D., exceptions being about as frequent relatively as in the older 
NT uncials. In its full development this flexion is characteristic of the 
second period of the Κοινή (i/A.D. and after), and is therefore only 
sporadic in LXX. Thackeray’s evidence there may prove that the infec- 
tion started in words with Ionic associations : this suggestion combines 
the alternatives discussed in Prol. l.c. It is difficult to say why the type 
ὀξεῖα did not conform to the new rule.? 

(b) Sreipa Lk 15° has no variant. It is the fem. of στεῖρος (usually of 
two terminations), and στείρα (sic scrib.) is a new fem.: see § 64. Note 
that the irreg. nom. ἑστηκυίη Wis 107 8* has no parallel in NT, and very 
few elsewhere. 

(ὁ) On the form Νύμφᾶν in Col 415 B, as an instance of levelling in an 
opposite direction, see Prol. 48. 

(d) Στοά (from στοιά) has a pure after the . has disappeared. 

(e) Πρύμνα is found in Ac 27*! without variant: in cl. Grk πρύμνη 
alternates with it. In μάμμη and θέρμα, where there is similar fluctua- 
tion, the NT exx. are indeterminate (gen. or dat.), while ἄκανθα (Attic) 
only occurs in plur.: see Thackeray 143, G. Meyer Gir.? 94. 


1 Prol. 38, 48. Cf. Thumb Hellen. 69 ff., WS 81 π., Thackeray Gr. 140-2, 
Mayser 12f., Schweizer Perg. 40ff., Kretschmer Hnistehung 30, Reinhold 48, 
Hatzidakis 84, CR xv. 34, 434 (papyrus evidence up to 1901), and xviii. 108. 

2 Blass and Thackeray would find a motive in Attic -ia for -via (Meisterhans 
59) which is assumed to make ἃ impure. But Attic had gen. -ὕας, and the 
power of cand p to influence a following @ had ceased to act centuries earlier. 
That ἀλήθεια οἱ sim. did not follow suit proves nothing, for nouns in -ia would 
supply a powerful analogy. 


§ 50-51] NOUNS. 119 


§ 50. B. Masculine Nouns in -as and -ης. 


veavias κριτής βορρᾶς ἄδης 
youth, judge. north (wind), Hades. 
Sing. N. ὃ νεανίας κριτής βορρᾶς ἄδης 
Vz νεανία κριτἅ βορρᾶ adn 
A. τὸν νεανίαν κριτήν βορρᾶν ἄδην 
G. τοῦ νεανίου κριτοῦ βορρᾶ ἄδου 
D. τῷ νεανίᾳ κριτῇ βορρᾷ adn 


Non-Greek proper names will be dealt with separately, 
Greek names in -@s and in -as impure follow normally the 
third of these models, those in -as pure the first. 


(a) Boppas is the usual Κοινή form, though the older Attic βορέας is 
rarely found. It is not a “contracted noun”—opéas—>opeas, as 
στερεός - oTepeds > στερρός. Meisterhans? 100 thinks the declension 
was adapted to the analogy of the “originally not Ionic-Attic abbrevi- 
ated names in -das.” Βορρᾶ gen. is the only NT form, but the whole 
of the above flexion is found in LXX (Thackeray 143). 

(0) “Αἰδὴ only 1 Co 1555 NeA? etc.: it is regular (K.BI. i. 387). The 
only vocatives found in NT are δέσποτα, ἐπιστάτα, καρδιογνῶστα, ὑποκριτά, 
Αἰνέα, ᾿Αγρίππα, and four Hebrew names (see § 60 for foreign names). 

(c) The gen. in -ov is specifically Attic (e.g. even Boppov), and naturally 
the alternative -a extends itself in the Kown, having a certain footing in 
Attic Volkssprache : in Lesbian (G. Meyer? 439) and in late Attic we even 
find exx. of gen. -y from nounsin -ns. (Cf. MGr κλέφτης, gen. -n.) It would 
probably have spread more if the Ionic had here agreed with the form 
characteristic of Greek outside Ion. Attic. Greek names in -as pure take 
του, a8 ᾿Ανδρέου, Λυσανίου (K.BI. 1. 386 f.}\—Josephus however shows many 
exceptions (Schmidt 489 f.); while those in -as and -as impure have -ἃ 
(-a)—thus Στεφανᾶ, ᾿Επαφρᾶ, and (6 conj.) ’Avrima (Rev 2!8—see Prol. 12). 
But usage differed for -as impure: cf. ᾿Αγρίππα P Amh 75 ter (ii/A.D.), but 
᾿Αγρίππον BGU ii. 511% * (ii/a.p., a copy of an official document of 
Claudius’ reign), and ᾿Ακύλα in BGU i. 71 (i/a.D.). See Schmidt Jos. 487 f. 
for similar fluctuations in Josephus: του here slightly predominates, but 
-a is common, and greatly outnumbers -ov in other names. According to 
Herodian (K.BI. 1. 386) marpod@as and μητρολῴας (as we spell in NT—see 
p. 83) had “ Doric” gen., but NT has only dat. pl. Cf. Thackeray 162 
on the “vulgar and late” use of -a in Hebrew proper names in LXX : 
also below, § 60 (4). 

(d) For cases of Metaplasmus see § 54. 


ὃ 51. C. Contracted Nouns. 
This category includes μνᾷ (like ἡμέρα) and γῆ, συκῆ, 
like φωνή. They differ only in that the accent is perispome- 


120 A GRAMMAR OF NEW TESTAMENT GREEK. [§ 51-52 


non throughout. The feminine of contracted adjectives 
(διπλοῦς, χρυσοῦς, ἀργυροῦς, σιδηροῦς) follows this model, 
with nom. δυπλῆ, χρυσῆ, ἀργυρᾶ, σιδηρᾶ. 


(a) Χρυσᾶν Rev 118 8*AC follows the analogy of ἀργυρᾶν, its natural 
associate. The pair react on each other in both ways—thus P Lond 12426 


χρυσῆν ἢ ἀργυρῆν. Blass (p. 25) gives a much less probable account of 
the genesis of this “ gross blunder,” as he calls it. 

(Ὁ) Uncontracted forms are occasionally found from χρυσῆ in Rev: 
χρυσέας 5° ᾿ξ, χρυσέων 21 AC. See below, § 64. 


SECOND DECLENSION. 
(1) Flexion. 


§ 52. A. Masculines and Feminines in -os, and Neuters 
in -ov. 


φίλος friend. 600s way. τέκνον child. 

Sing. N. ὃ φίλος 7 ὅδδός τὸ τέκνον 
γ. φίλε τέκνον 

Α. τὸν φίλον τὴν ὁδόν τὸ τέκνον 

G. τοῦ φίλου τῆς ὁδοῦ τοῦ τέκνου 

D. τῷ φίλῳ τῇ ὁδῷ τῷ τέκνῳ 
Plur, Ν. οἱ. φίλοι αἱ ὃδοί τὰ τέκνα 
Vz. φίλοι τέκνα 

Α.. τοὺς φίλους τὰς ὁδούς τὰ τέκνα 

G. τῶν φίλων τῶν ὁδῶν τῶν τέκνων 

D. τοῖς φίλοις ταῖς ὁδοῖς τοῖς τέκνοις 


(a) Nearly thirty vocatives in -e occur in NT: this formation 
survives in MGr. No feminine -os nouns in the NT show the vocative. 
Θεός makes θεέ nearly always in LXX, as in Hellenistic generally (θεός 
Att.): ef. Thackeray 145, CR xv. 34, 434. Τιμόθεε (Lucian) answers to 
old Attic models. Note υἱός voc. once in Mt (12°—vié appears four 
times): cf. Mayser 256. 


B. Contracted Nouns. 


The norm, may be seen in the masc. and neut. of 
adjectives: thus 


Sing. NV. διπλοῦς διπλοῦν Plur. διπλοῖ διπλᾶ 
Α. διπλοῦν διπλοῦς διπλᾶ 
G. διπλοῦ διπλῶν 


1). διπλῷ διπλοῖς 


§ 52-53] NOUNS. 91 


(a) For νοῦς and πλοῦς, originally in this class, see § 59 (4). 

(b) Open forms, presumably Ionic in origin (Thumb Hellen. 63), are 
found rather freely in the flexion of χρυσοῦς in Rev: thus 2! χρυσέων AC, 
4: -gous &, 58 -éas δὲ, 939 -αἰα δὲ, 1414 -eov (in 38 only); so χάλκεα 929 XN, 
Thumb’s statement that they were characteristic of the Eastern Κοινή is 
suggestive in connexion with the curious fact that they are peculiar to 
Rey. On the other hand the flexion of ὀστοῦν, the only substantive in 
this class in NT, shows open forms in the plural: ὀστέα Lk 3439 (-a@ DN), 
ὀστέων Mt 2377 Heb 112? (Eph 59° ΝΟ) etc.). This differs from LXX, 
where the rule is that uncontracted forms come in gen. and dat. sing. and 
plur. (Thackeray 144: see also 172 f. on the adjectives). See CR xv. 35, 
435 ; Schmidt Jos. 490 ff.; K.Bl. i. 402 (§ 113 n. 3). 

(c) Xequdppov in Jn 18! may probably be accented thus (so Blass 25), 
as coming from a late shortened form χείμαρρος ; but χειμάρρους occurs 
normally in LXX (Thackeray 144), 


C. “Attic” Declension. 


Strictly this declension affects a few words which by 
“metathesis of quantity ” had substituted -ws for the final -os. 
Thus νεώς, λεώς from νηός, Ads, for which Hellenistic replaced 
(except in the compound νεωκόρος) the general Greek vads, λαός. 


With a different history we have ἅλως (ὁ) threshing-floor, which still 
survives in LXX and papyri, though replaced in NT by ἅλων (4): its 
flexion (sing.) is A. ἅλων, G. ἅλω, 1). ἅλῳ. In Ptolemaic papyri (Mayser 
259) A. plur. ἅλω(ς), G. ἅλων. See in general Meisterhans® 129-131. 

(a) For the neuter ἀνώγεων Ti cites 15 cursives in Mk 14!) and names 
two (with “al”) in Lk 2212 See WS 47. It has no classical or 
Hellenistic warrant. 

(b) Ἵλεως survives only in the nom. sing. It had been stereotyped 
largely by the phrase ἵλεώς σοι (etc.) “mercy on thee!”: see Prol. 240. 
It was rather more alive in the LX X period—see Thackeray 173. 

(ὁ) ᾿Απολλώς follows this model in N.A.G., with the same fluctuation 
as in Attic (Goodwin Gram. ὃ 199) between -ών and -ὦ in acc. Thus 
1 Co 4° ᾿Απολλών N*¥AB* but ᾿Απολλώ Ac 19! exc. A?L 40, See Meister- 
hans® ὃ 49d. There was hardly any difference in sound. 

K@s in Ac 21! has ace, Κῶ exc. in HLP. 

Both these nouns have been preserved by the influence of the mixed 
declension. 


(2) Gender. 


ὃ 53. (a) Neuter plurals attached to a mase. sing. in -os 
are still found in NT. They were originally collectives and 
some traces of this sense survive. 

Acopés has plur. δεσμά in Lk 8” Ac 1056 205", δεσμοί 

1 Giles? 266 ff. 


122 A GRAMMAR OF NEW TESTAMENT GREEK. [8 53 


Phil 11%, the rest being ambiguous. Thackeray (p. 154) 
observes that in LXX as in NT δεσμά is literary. 


Rutherford Gram. 9, asserts after Cobet that Secpa=actual bonds, 
Seopoi=bondage. The distinction cannot be pressed for the NT, though 
it would suit very well: Ac 207° gains vividness from it. The original 
differentia is in this case almost inverted. 


Θεμέλιος (sc. λίθος) has masculine forms except in Ac 1676, 
Ta θεμέλια here shows the collective sense: contrast of 6. in 
Rev 217°. 


It is common in LXX, where the masc, appears rarely. Thackeray 
(154) suggests that the earlier and later Κοινή levelled in different 
directions, “the former using the neuter throughout, the latter the masc.” 
Thucydides however has οἱ 6. (i. 93) and so has Aristotle. Moeris and 
Thomas Magister declared the neuter alone Attic: it may be questioned 
whether they had adequate data, See Mayser 289, Cronert 175. 


Xiros shows the old plur. ofra in Ac 7! HP, where Blass 
says σιτία does not suit; but see Wendt or Knowling. It 
survives in two literary LXX books (Thackeray 155), 

Στάδιον retains the old double plural, but στάδιοι. pre- 
dominates: στάδια only Jn 61° 8*D 106 (against x*® ABL 
etc.). 

Thackeray assigns στάδιοι to the literary element, which NT use 


makes doubtful. Kalker (p. 239 f.) says that Polybius alternates the 
forms to avoid hiatus, which makes them mere equivalents, 


Λύχνος and χαλινός have only masc. plur., as in LXX. 


(Ὁ) Gender fluctuates in the following :— 
ἤλβυσσος, an adj., becomes a fem. noun (80. χώρα). 


First in LXX (as Gn 12), but a citation from Diogenes Laertius 
(ii/A.D.) shows it was a “ profane” use: see Grimm-Thayer s.v, 


᾿Αλάβαστρος is fem. in Mk 14% 8°BCLZA, mase. in 8*AD 
etc., and actually stands as neuter in GM 1 13-69 (Ferrar). 


LS quote for the neuter Theocritus and the Anthology. The Attic 
ὁ ἀλάβαστος accounts for the second alternative, which occurs in LXX 
(4 Καὶ 2118 B). Ti quotes a scholiast who writes τὴν d., observing that 
Herodotus made it mase. Blass (p. 26) says Attic should be ἡ 4., though 
he quotes no authority and admits 6 ἀλάβαστος for Aristophanes and τὸ 
ἀλάβαστον for Menander. The word probably comes from Arabic (see 
Boisaeq s.v.), and naturally wavered when a gender had to be found, 
much as German wavers between der, die and das Awesta. 


§ 53] NOUNS. 123 


“Aupodoy neut. in Mk 114 as usual, but see Mayser 261 n. 
ἤλψινθος in Rev 8" is ὁ ἄ,, but x* substitutes the more 
usual ἀψίνθιον. 


Fem. in Aretaeus (medical—i/a.p.). From a pre-Greek place-name, 
according to Kretschmer: Boisacq gives ᾿Αψίνθιοι as a Thracian tribe. 
WS 83 accounts for masc. in Rev 811 by noting it is an angel’s name, 


Βάτος is fem. according to Moeris in Hellenistic. So in 
Luke’s use: see Blass on Ac 785, 


Thackeray (p. 145) appears to be wrong in making the LXX mase, 
the Κοινή norm, unless Moeris is the blunderer. Mk 1236 has masc. (2). 
We might read here (with RV and Swete) ἐπὶ τοῦ Barov from τὸ “ Bdros,” 
“the Bush passage.” SH on Rom 11? show that ἐν without article is 
normal in this locution, but their evidence is hardly decisive. The order 
of the words in Mk favours the RV translation: Luke’s change of order 
and gender may well mean that he took it locally—it was actually “at 
the Bush” that Moses made the pronouncement. 


Δεῖπνον appears as masc. in Lk 14:6 ΒΡ 4Π5", Rev 19° 
046 cut, 1917 cu?°: so MGr. 

Zuyés is regular in the Kowy since Polybius for ζυγόν : 
earlier masc. only in sense of balance (see LS). 

Anvés has (class.) fem. in Rev 14? dis 1915: but in 1419 
τὴν Δ... .. τὸν μέγαν (x corrects)! It is only another 
instance of the breach of concord familiar in Rev. See Swete, 
also Charles in loc. 

Ὁ Anvos is given by LS from Athenaeus xi. 49 (p. 474 fin.), and Is 632 


(probably wrong—see Ottley in loc.). Thackeray quotes cursives of 
Gn 8088, 41, 


Λιβανωτός 15 written -ov, neut., in a few cursives at Rev. 8°. 
Ai®os is no longer fem. when meaning gem (Rev 21" etc.): 
so LXX. 


Meisterhans 129 says the fem. is frequent from 385 B.c. If it is 
specifically Attic, and late at that, it would naturally get no footing in 
the Kown. Cf. ὕαλος below, and see Mayser 262. 


Λιμός Wavers In gender: the fem. was a Doric element in 
the Kown (Thumb Hellen. 67) and as such unstable. 


Phrynichus says τὴν λιμὸν Δωριεῖς ; Lobeck (p. 188) supports it 
convincingly. Moeris calls the fem. Hellenistic : Mayser (p. 8) emphasises 
this with evidence. It is sporadic in LXX (Thackeray 146). NT in- 
stances are indeterminate except in Luke: fem. is certain in Lk 1514 
NABDL, Ac 1178 ΝΣ ΒΞ; but in Lk 455 only 13-69 (Ferrar) evidences 


124 A GRAMMAR OF NEW TESTAMENT GREEK. [§ 53-54 


μεγάλη. See Prol. 60 for a case of wavering in papyri written by the 
same hand, which excuses our seeking an explanation in varying sources 
used by Luke. 


Νῶτος in Rom 111° (LXX) replaces classical νῶτον. 
So in LXX, exc. Gen 9558 Jer 277 (Thackeray 155). 


Σάρδιον is masc. in a good many cursives in Rev 21”, 
Στάμνος keeps fem. gender in Heb 95: masc. in Ex 16% 
(exe. one cursive). 


Mayser (p. 262) cites three papyri (Ptolemaic) for ὁ o., and notes that 
Sextus Empiricus gives the masc. as Peloponnesian, fem. as Attic. 


TpiBos (only in a LXX citation) keeps fem. without 
variant. 


It wavers very much in LXX (Thackeray 146). 


Ὕελος (or vados—see ὃ 33. 1) fem. in Hdt. etc., mase. in 
Rey. 21%. 


LS cite Theophrastus for the masc. 


Ὕσσωπος indeterminate in NT: see Thackeray 146. 


Metaplasmus and Heteroclisis. 


§ 54. Fluctuations of gender within the Second Declension 
have been already dealt with: we present here only Greek 
nouns which appear in various declensions. Proper names 
and foreign nouns are treated separately. 


A. Declensions I. and II. 


-άρχης and -apxos compounds. The Decl. I. form has 
encroached very largely on the Attic -apyos. In NT only 
one word keeps -apyos without wavering, viz. χιλίαρχος (as 
LXX). ‘Exatovtapyos is overwhelmingly predominant in 
LXX, but WH accept it only 4 times out of 17 places 
(apart from ambiguous gen. sing. and plur.). 2 τρατοπέδαρ- 
χοὸς appears in HLP al. at Ac 2816, but the clause (accepted 
by Blass for his 8-text) has slender authority. No variants 
occur for ἐθνάρχης, πατριάρχης, πολιτάρχης and τετραάρχης. 
᾿Ασιάρχης only occurs in gen. plur. (Ac 1951), but we should 
probably accentuate -ῶν (cf. -ynv in JMAe ili. 525, 526— 
Thera, ? ii/A.D.). 


$54) NOUNS. 125 


‘Exardvrapxos (excluding gen.) occurs in the great uncials in only five 
places viz. δὲ ὅς, By, Ayo, D#: WH give in Mt. nom. -os dat. -y, in 
Luke -ns throughout, but acc. -ov. Such mixture is paralleled in 
papyri: see CR xv. 34, 434, xviii. 108, Mayser 256f., where literature 
on the subject is given—add Thackeray 156. Mayser observes that 
-dpxns started in Ionic districts (so Herodotus), spread thence into Attic 
tragedy, and prevailed more and more in the Kowy. New formations 
were almost exclusively of this form, while in compounds of numerals 
and old official titles the other kept its own until the Ptolemaic period 
was past. Thumb (Hellen. 59) calls attention to the significant fact that 
in Attica -apyos persisted until A.D. (middle of i/a.p., to judge from exx. 
in Meisterhans* 125): this will then be an element in the Kou} 
decidedly traceable to non-Attic influences, 


Δυσεντέριον in Ac 28° xAB αἱ. for -ia, was Hellenistic 
according to Moeris. 

Ἔνεδρον (for ἐνέδρα) occurs in LXX (Thackeray 156), but 
not in NT (Ac 23" only HLP). 7 

*Hxos (masc.) replaces ny7 from Aristotle down. So 
Heb 12, and other places where it might be Decl. III. (see 
below, B (a)). 

Θεά seems to have been the Kown fem. of θεός: in 
Ac 195] we find τὴν θεόν used as the regular term. techn. for 
the city goddess—see Prol. 244. 


Blass held that ἡ θεός was Hellenistic, except in the formula ἡ μεγάλη 
or ἡ μεγίστη θεά : inscriptional evidence from Magnesia strongly supports | 
the other (Kuhring’s) view. Ἢ θεά appeared in Attica itself in iii/B.c. 
(and earlier in direct antithesis to ὁ θεός) : see Meisterhans* 125. 


B. Declensions I. or 11. and III. 


(a) Neuters in -os have increased at the expense of the 
first and second declensions. This arises naturally from the 
coincidence of nom. sing. in the masc. and neut. -os nouns, 
and their nearness in gen. sing, where an -s was easily 
added or lost. Similarly even in classical times there was 
confusion between nouns in -ns gen. -ov (Decl. I.) and those 
in -ys gen. -ovs (Decl. III.)—cf. G. Meyer Gram? 439 f. 
The confusion has developed in MGr: Thumb Handb. 64. 

Δίψος and δίψα both occurred in Attic (Blass 28). NT 
only 2 Co 1151, where B* has δίψῃ, and the rest δίψει. 


P Flor ii. 1761? (Α.Ὁ. 256), an illiterate private letter, has gen. δίψης, 
while the medical fragment P Tebt 11. 27217 (late 1i/a.p.) has τὸ δίψος. 


126 A GRAMMAR OF NEW TESTAMENT GREEK. [8 54 


Ἔλεος as neuter always in NT, nearly always in LXX. 


See Thackeray’s interesting evidence (p. 158) as to the literary 
character of ὁ €. (Attic) in its few LXX occurrences. (Add that it 
alternates with τὸ ἔ. in Pss. Sol). Adjectives like ἐλεεινός and σκοτεινός 
suggest that the neuter forms in these words were survivals. But to 
prove this we must postulate their coming into the Kowy through some 
dialect that preserved the hypothetical old neuter ; Brugmann (Grd.? τι, 
i, 282) treats them as analogical extensions from φαεινός (= φαεσ-νός) and 
the like. The adj. νηλεής however goes the other way. [Ἔλεος masc. 
only survives in later uncials, with one appearance in C (Mt 237%), 


Ἕλκος (τό) has acc. sing. ἕλκον in Rev 16? x*. 

Ζῆλος is neuter in Ac 5” B*, 2 Co 9? xB 33, Phil 3° 
x*ABD*FG: ὁ €. occurs in seven places (eight, if we followed 
NCD°w in Gal 5°), 


Ὃ ¢. rarely occurred in LXX. It is neuter in ΜΟΥ, which makes for 
the view (WS 84) that the neuter was popular Greek in Hellenistic 
times. 


*Hxos is of Decl. II. in Heb 121%, of III. in Lk 21% 
(neut. or fem.). 


In Lk l.c. WH (App.? 165) accentuate ἠχοῦς from ἠχώ f., Ln and Ti 
ἤχους from τὸ ἦχος. The existence of the latter is proved from LXX 
(Thackeray 159), and from several quotations collected by Schmiedel 
(WS 84 n.), who remarks that the meaning sound for ἠχώ is only poetical, 
except in Philo i. 588 and Job 4:3 (see below). Whether ἠχώ survived 
in vernacular Greek can hardly be determined. Its influence may 
perhaps be traced in Job 41%, where ἠχῷ is fem.: should we accent ἠχῷ 
with » for o! (see § 36)? In Lk perhaps ἤχους is slightly more 
probable : the OT original (Ps 657) to which WH assign it has ἤχους acc. 
pL—gen. sing. is barely possible. (The ancient conjecture (?) ἠχούσης, 
found in D and Eusebius, would improve the construction.) Ἦ χος is 
mase. in MGr. 


Θάμβος is neuter in NT, but θάμβον gen. in Ac 3 C: 8. 
μέγας Lk 45 Ὁ. 


It is mase. and neut. in classical Greek (Blass 28) and LXX 
(Thackeray 158). 


Νῖκος (τό) has supplanted νίκη in true Hellenistic, though 


the latter survives in 1 Jn 54. To v. in Mt 1279, 1 Co 
1554: 55. 57. 


Neut. in BGU iv. 1002!4 (55 B.c.). The old fem. is literary in LXX 


1 Of. Λητῷ dat. in a Phrygian inser. (JHS iv. 385) ap. Dieterich Unters. 
163. 


§ 54] NOUNS. 127 


(Thackeray 157): cf. P Lond 1178!? (=iii. p. 216), where it is used in a 
letter of Claudius. See Mayser 93 n.8 for exx. 


Πλοῦτος is neut. in nom. and acc. 8 times in Paul: masc. 
(nom. ace. gen.) 5 times, and 7 times in other NT writers. 


Neut. only once in LXX, Is 29? (but ὁ 7. BQ). It is MGr. 


Σκότος, formerly masc. and neut., is always neut. in LXX 
and NT. 


The gloss σκότῳ has intruded in late authorities at Heb 1218, 


Στρῆνος (τό) has gen. στρήνου in Rev 188 C cu’, 

(6) Contracted masc. nouns of Decl. II. (substantives 
only) have passed into Decl. III, with which they already 
coincided in nom. and ace. sing. (Bods, βοῦν). Thus νοῦς 
makes gen. νοός, dat. νοΐ, πλοῦς gen. πλοός. 


So even the Atticising writer of 4 Macc (Thackeray 160). To the 
inscriptional exx. in WS 84 n.’ add peas from ῥοῦς P Oxy iv. 736°8 
(c. A.D.1). But νόῳ BGU 1]. 385° (ii/iii a.D., an illit. letter), and Εὔπλοος 
REGr xvii. 205f. (Rhodes, ii/a.p.). See further evidence in K.BI. 
i. 516, G. Meyer 419, Crénert 166. The plural nom. εὔνους is found 
in Attica as early as 300 B.c.: the analogy given above could not have 
produced this, which is presumably an accus. influencing nom. Χοῦς 
(liquid measure) was declined like βοῦς even in Attic, and this may well 
have affected the other χοῦς, “earth” (χοός, χοΐ in LXX) so starting the 
type: cf. χοΐ in IM Ae iii, 248 (ii/B.c.). See Mayser 257, whose papyrus 
evidence shows that the type had not developed far in the first (B.c.) 
period of the Kon, 


(c) Miscellaneous instances under this heading are— 
ἍΛλων (77, gen. ἅλωνος) has. replaced ἅλως (see above, ὃ 53). 
Thayer cites it from Aristotle. In papyri it occurs, but far less often 


than ἅλως : see Mayser 258 f., 287, and add the early instance P Lille 13° 
(243 B.c.). 


Γόης makes pl. after Decl. I. γόηται in 2 Ti 313 D*, 

Adxpuov has the dat. pl. δάκρυσιν Lk 1738. 44 a survival 
from the old δάκρυ, which agrees with δάκρυον in nom. acc. 
gen. pl. See below, p. 141. 

Κατήγωρ in Rev 1210 A is said (WS 85) to be only the 
Aramaic term 0p, a Greek loan sent back in damaged 
condition. But cf. Thumb Hellen. 126. 


Blass calls in the analogy of ῥήτωρ : the two types coincide in gen. 
pl. Schmiedel compares the late forms διάκων for διάκονος and πάτρων 


128 A GRAMMAR OF NEW TESTAMENT GREEK. [8 54-55 


for Latin patronus (a loan-word). But this is really an alternative to the 
explanation quoted from him above: these words, for which no foreign 
borrowing can be appealed to, must be explained by a mixture of 
declensions characteristic of the later periods of the Kowy. This is 
Thumb’s view, 


Σάββατον was a Semitic word, and should properly come 
in below (p. 153). But its dat. pl. σάββασιν (once in LXX, 
always in NT) is on a good Kown model: cf. Lobeck Par. i. 
175, where grammarians are cited showing that πρόβατον 
made πρόβασιν in dat. pl—this case does not appear in NT. 


WS 85. π.8 cites σάββασιν from Jos. Ant. xvi. 6 and Meleager 834. 
W. F. Moulton (WM 73 n.1) gives σαββάτοις (the LXX form) from 
Mt 12!}12.B. See Schmidt Jos. 499 f. 


(ad) Heteroclisis in proper names (non-Semitic) may be 
noted in the following :— 

Θυάτειρα is neuter pl. except in Rev 11 AC 046 Θυάτειραν 
acc. (-a δὶ and so WH text), 2% -ρῃ dat. δ} vg, and even 
-pats dat. pl. in late MSS. 

Λύστρα similarly has acc. -av, dat. -οἰς in the same 
context: see Prol. 48, § 60 (10) below. 

Muppa is neut. pl. in Ac 27°, but 81 reads Mupav, which 
Ramsay supports from the modern name: the gen. is Μύρων 
(or with pp). 

᾿Σαλαμίς (? nom. Σαλαμίν on analogy) has Σ᾽ αλαμίνῃ as “a 
well attested substitute for” its regular dative (WH App. 
163): so in Ac 13° 8AEL and some Latin texts — cf. 
Reinhold 56 for late vernacular evidence. Suidas (p. 413a 
Bekker) gives Σαλαμίνη (in gen.) as the older name of 
Constantia in Cyprus: WS 94 also cites Salamina-ae from 
Justinus for the more famous Salamis near Athens. 


THIRD DECLENSION. 


§ 55. In this Declension are grouped together a great 
variety of stems, all ending in consonants or semivowels (4 or 
u). ΑΒ the semivowels and o fall out between vowels, a good 
deal of contraction results. A marked feature of the declen- 
sion is the (very limited) survival of “strong flexion,” by 
which a stem varies through vowel-gradation (Ablaut) in 


§ 55] NOUNS. 129 


different cases. The proper case-terminations will be seen 
in the types of nouns with consonant stems which do not 
contract, nor show strong flexion. 
A, Stems ending in Mutes. 
σαρκ- (ἡ) μαστῖγ- (ἣ) πνευματ- (τό) παιδ- (ὃ, ἡ) 


Flesh. scourge. spirit. boy, girl. 
Sing. N. σάρξ μάσττξ πνεῦμα παῖς 
Α. σάρκα μάστιγα πνεῦμα παῖδα 
G. σαρκός μάστιγος πνεύματος παιδός 
D. σαρκί μάστιγι πνεύματι παιδί 
Plur. N. σάρκες μάστιγες πνεύματα παῖδες 
A. σάρκας μάστιγας πνεύματα παῖδας 
σ΄. σαρκῶν μαστίγων πνευμάτων παίδων 
D. σαρξί(ν) μάστϊξι(ν) πνεύμασι(ν) παισί(ν) 


N.B:—(1) Vocatives will be specially mentioned in this declension 
whenever separate forms occur. There are none here except γύναι : the 
classical παῖ is obsolete. 

(2) Monosyllables transfer the accent to the last syllable in gen. and 
dat.: παίδων is an exception. 

(3) The acc. sing. in -ay is conspicuous in the vernacular throughout 
the 3rd decl., but the culture level of the documents that exhibit it is 
decidedly lower than anything we find in NT. It does not appear in 
Attic inscriptions till the later Roman period :! see Meisterhans® § 50. 7. 
For its currency generally cf. Schweizer Perg. 156f., Schmid Attic. 
iv. 586, Cronert 169 (and reff. there), Jannaris pp. 542f., CR xv. 34f., 
435. We will deal separately with the rather different case of -ν added 
to acc. -n or -ἢ : see below, ὃ 58 (d). In LXX (Thackeray 22) the pheno- 
menon is almost confined to δὲ and A; nor is the case very different in 
NT, for Scrivener (Collation p. liv) cites ten exx. from δὲ in NT and 
fourteen more from Barnabas and Hermas, while A has at least five. 
The following list is perhaps sufficiently complete. Mt 2? dorépav X* 
(for ἀ. ἐν), 21° ἀστέραν N*C, 53° τρίχαν N* EL etc., 918. χεῖραν L, 1249 χεῖραν 
ΝᾺ 80 far of course A Aiat. Mt 2775 χλαμύδαν D, Mk 1*! χεῖραν A*, 657 
σπεκολάτοραν D*, 130 θυγατέραν D, 7°? yeipay D, Jn 6°4 σάρκαν D, 207° 
χεῖραν N*AB, Ac 6° ᾿Αντιοχέαν Οὐ, 14! Δίαν DE al., 16° Τρωάδαν 61, 17° 
Ἰάσωναν D*, 217 Πτολεμαΐδαν δ δ, 2275 ἀέραν &, 1 Pet 5° χεῖραν NA, Heb 8° 
δειχθένταν DE (107! ἱερέαν L—but here there is -a final, which brings the 
case near to those in ὃ 58 (d) below), Rev 6° and 9* odpayiday 8 and a 
cursive or two, 1218 ἄρσεναν A, 1314 εἰκόναν A, 22? μῆναν A. On the case 
as it affects the NT, WH (App.? 164) pronounce generally that the ν is 


1 Δήμητραν is printed in the Teubner text of Plato Cratylus 404}, but 
silently emended in the Oxford text: its MS attestation can hardly be regarded 
as evidence here. 


130 A GRAMMAR OF NEW TESTAMENT GREEK. [§ 55 


due to transcribers, both where added to -ἅ and in the contracted stems 
(p. 189). They are influenced by “the irregularity and apparent capri- 
ciousness of its occurrence,” the generally scanty witness and especially 
“its extreme rarity in B.” To this we may add the curious fact that it 
appears most often (except for Rev) in the writings where the Greek is 
good, while the papyri show it characteristically in documents of low 
culture. In later periods of the Kowy the levelling process brought the 
final -ν everywhere into the acc. sing. in popular speech, to disappear 
again impartially except in certain dialects and under certain sandhi 
conditions. Thus in ΜΟῚ we find τὴμ πίστι, τὴ μέρα, τὴν ἐρπίδα (πίστιν, 
ἡμέραν, ἐλπίδα) : see Thumb Hdb. ὃ 34. 

(4) Acc. pl. m.f. in -es, encouraged by the identity of nom. and ace. in 
such flexions as πόλις, γραμματεύς, πῆχυς, is fairly common in papyri of 
early and later periods. It was probably started by τέσσαρες, which in 
LXX and papyri is far the commonest instance (Thackeray 148): in 
some kinds of writing it outnumbers τέσσαρας ; see Prol. 243 f., also 36, 
where it is noted that there is good uncial authority for -es in every NT 
occurrence of the accus. That δύο, τρεῖς, πέντε κτὰ. have no separate 
accus. form is enough to account for this form. Apart from this there 
are no NT exx. 


(1). Guttural Stems. 


᾿Αλωπεκ- (ἡ) fox, pl. ἀλώπεκες, has nom. sing. ἀλώπηξ. 
For gender see under SYNTAX (Vol. 1Π|.). 

Γυναικ- (ἡ) woman makes voc. γύναι (with final « dropped), 
and takes for nom. sing. an old Ist decl. stem. γυνή (cf. 
Gothic quind). Its accent follows the monosyllables— 
γυναῖκα, γυναικός, etc. 

Θριχ- (ἡ) hair is affected by the law which forbids suc- 
cessive aspirates:1.N. θρίξ, A. tpixa(v) (see above) D.PI. θριξί(ν) 
ete. 
Κηρῦκ- (ὁ) herald, like Φοινῖκ- (ὁ) Phoenician Φηλικ- (0) 
Felix and χοινῖκ- (ἡ) quart, is variously accented in nom. 
sing. κῆρυξ (WH) and κήρυξ (Ti) according as we accept or 
reject express statements of ancient grammarians: see § 29, 
and especially K.Bl. 1. 420. 

ὌὈρνῖχ- (ἡ) hen, nom. sing. ὄρνιξ Lk 13° 8D. “Opus (¢.v.) 
is the reading of WH, but the rarer form has a strong claim. 


ἼὌρνιξι occurs six times in P Lond 131 (i/A.D.) (=i. 109-88). It was 
mentioned Prol. 45. An element drawn from Doric, it was probably 


[᾿ Aspirated tenues lost the aspirate in primitive Greek when the next 
syllable or next but one also began with an aspirate. Brugmann ΟὟ, 122.— 
Ep.] 


§ 55] NOUNS. 131 


dialectic in the Kown, just as its descendant ὀρνίχ is local (Cappadocian) 
to-day. Photius (ap. Ti) says it was also used by the Ionians, with the 
oblique cases. This suits its modern survival very well. Crénert’s 
instances of it (p. 174n.) should be observed. 


(2) Labial Stems. 


Four nouns in 7- and one in β- occur in NT. The nom. 
of course is in -~: there is nothing irregular or noteworthy. 


(3) Dental Stems. 


(a) Very many nouns make stem in -6. Those in -ds gen. 
-ἄδος, -is gen. -(dos, -ς gen. -vdos, keep the accent on this 
syllable throughout. In odparyid- (ἡ) seal the t makes all 
oblique cases (exc. gen. pl.) and the nom. pl. properispomenon 
(σφραγῖδα etc.). 

Svyyevid- (ἡ) kinswoman, nom. συγγενίς, serves as a Hel- 
lenistic fem. for συγγενής. By earlier rule only nouns in 
-τῆς cf. (προφῆτις, πρεσβῦτις) and -evs could form such fem. 
If the rule is to be maintained for later Greek, we might 
observe that συγγενής has sundry forms from the -evs flexion 
(see § 59 (3)). 

Ποὃ- (0) foot keeps its irregular nom. πούς : the rest is 
normal. 

(Ὁ) Barytones in -ἐς (gen. -etos and -cdos), being brought 
by their accent into asssociation with the -ez stems (§ 59 (1)) 
had in earlier Greek ace. sing. in -ἰν. So in NT προφῆτιν 
from προφητιδ- (ἡ) prophetess. More or less stem-mixture 
appears in the following :— 


Ἐριδ- (ἡ) strife. Ace. sing. ἔριν only, but also in pl. we find all the 
versions (exc. Eth.) supporting ἔρεις in Tit 3° (so Ne AC). WH reject it 
with 8*D*G, one singular being easily assimilated to the plurals around. 
Where ἔρεις nom. pl. occurs, there is always a variant ἔρις, and we cannot 
feel any confidence in it. WH place it in margin “with hesitation” at 
Gal 57°, But when εἰ and « were identical in pronunciation it is unlikely 
that such a new form would oust the regular ἔριδες (1 Co 11!—no v,/.), 
and produce a needless ambiguity. 

Κλειδ- (ἡ) key was not originally a ὃ- stem: οἵ, Lat. clavis—the gen. 
kAn(F)idos is however as old as Homer. It kept in Attic its proper ace. 
κλεῖν, pl. κλείς (K.BI. i. 461): so in NT Rev 37 20! and 118 respectively, 
without serious variant, also Lk 115? D, Mt 1019 8¢B2CD al. The LXX 
forms κλεῖδα κλεῖδας (Thackeray 150) stand in Mt 16!9 8*B*L, Lk 1152 
(exe D), and are introduced in cursives (exe Mt lc.) : κλεῖδα is expressly 

VOL ἘΠ PART 11:-- )]0 ; 


132 A GRAMMAR OF NEW TESTAMENT GREEK. [8 55 


said by Moeris to be Hellenistic, and it is also banned by Phrynichus, 
Thomas and the Antiatticista. But it is odd that Attic κλεῖν and κλεῖδα 
should appear ‘as well as κλεῖδα -as in the papyri,! and odder still to find 
the author of Rev. among the Atticists, with the other NT writers using 
correct Hellenistic ! 

Νηστιδ- (ὁ or ἡ) fasting, which strictly should be called an adj., shows 
in classical Greek the alternative stems νηστῖ- and νηστει--. Since the 
-7- stems are obsolete in Hellenistic, we may safely reject vnoris as 
nom. sing. in Dn 618 or as ace, pl. in Mt 1653 Mk 8°: it is only Itacism for 
νήστεις in the latter and νῆστις in the LXX of Dn lc. The stem then 
may perhaps be set down best as νηστει- ; but Phrynichus (Lobeck 326) 
seems to imply that the Hellenistic word was νήστης : cf. Syll. 8059 
(2i/a.D.), and the medical papyrus P Oxy viii. 1088** (early i/a.p.). Was 
the word heteroclite, with 1st decl. forms in the sing.? See WH App.? 164. 

Χαριτ- (7) grace keeps Attic acc. χάριν some forty times, but has 
χάριτα Ac 2427 R*¥ABC, 25° A, Jude* AB, which according to Moeris 213 
was Hellenistic. It is well supported in the vernacular of the imperial 
age; see CR xv. 35; Thackeray 150; Mayser 271f. and reff. there. 


(c) Stems in -7 are mainly accounted for by the large class 
of neuters in -uat-,and by the fem. abstracts in -ry7-. In the 
former the stem in -mm- has been extended by fusion with 
words in -muto-. There is a small class of neuters in which 
hysterogenous -7- forms have from an early period ousted 
more original ones from -ασ- base (see ὃ 58 (3)): 80 κέρας horn, 
τὰ κέρατα, πέρας end, τὰ πέρᾶτα, τέρας marvel, τὰ τέρᾶτας The 
types coincide in 1). plur. τέρασιν. Cf. Meisterhans* 143. 


“Adas (τό) salt, gen. ἅλατος, has largely driven out the older ἅλς (6). 
In Ptolemaic papyri and LXX ads predominates (Mayser 286, Thackeray 
152) ; but ἅλας is certain in 2 Esd 7?? Sir 39°62 In NT as disappears, 
except for ἁλί Mk 94 D (from LXX) and ἅλα Mk 9580 ace. But this 
last may belong to the variant nom, ἅλα (Lk 1455 bis 8*D, Mt 518 Nbis 
DWhis, Mk 95° LAbis N*semel), which appears also in Sir 3936 NBC 
(ἅλας A). In Lev 213 we find ἅλα and ἅλας in the same verse translating 
the same Hebrew,® which starts a possible hint for the genesis of these 
forms. Im 14 LXX occurrences of ἅλα and ἅλας accus. the article is 
absent, and there is nothing to show gender or number, if it were not for 


1 Mayser 272, CR xv. 35—add P Oxy iii. 50234 (A.D. 164) θύρας καὶ κλεῖς. 

2 In a letter dated Jan. 10,1911, Mr. Thackeray agrees with this statement, 
and adds that he would now regard ‘‘the other exx. of ἅλας (and perhaps 
dda)” as ‘‘ probably neuters. The only indubitable cases of the plural are in 
the local plural phrases ἡ θάλασσα (κοιλάς, φάραγξ) τῶν ἁλῶν. This looks as if 
the plur. was the regular form for salt-areas (salt-marshes ete.) in which the indi- 
vidual lumps or particles were widely distributed.” 

3 There are variants both ways, but not in the major MSS: see the Cam- 
bridge LXX «in Joc. 


§ 55] NOUNS. 133 


17 places where dds, ἁλί and ἁλῶν appear. Very possibly the new noun 
arose from these ambiguous forms ; it may have been encouraged by the 
formal similarity of two other food-names naturally associated with it, 
γάλα and μέλι, which were both neuter. According to ancient gram- 
marians (K.BI. i. 456) yada as well as μέλι had a gen. formed by simply 
adding -ros, which may have helped the new flexion. The old ἁλός can 
be cited as late as iii/a.p., in P Lond 1170 vs. 124 (=iii. p. 196): ef. WS 
90 n. 

Tadaxr- (τό) milk has ace. γάλα, gen. γάλακτος. This and νυκτ- (ἡ) 
night, nom. νύξ, are the only stems in which τ follows a mute. 

Τελωτ- (6) laughter only occurs in nom. γέλως, as does ἱδρωτ- (6) sweat, 
nom. ἱδρώς. For earlier flexion see Κὶ.ΒΙ. i. 516 and 509f. 

Ἔσθητ- (ἡ) clothing has ace ἐσθῆτα etc., but a heteroclite dat. pl. 
ἐσθήσεσι in Lk 244 (all exc. NBD, which have ἐσθῆτι), Ac 110 NABC: 
Deissmann supports this with BGU i. 1613 (159 a.D., as amended, p. 395), 
and Crénert (p. 173) adds several citations from MSS of Κοινή writers. 

Φωτ- (τό) light has nom. ace. φῶς, and χρωτ- (6) skin, gen. χρωτός, nom. 
χρώς. Both have considerable variation of stem in the earlier Greek : 
see K.Bl. i. 436,511. φῶς is accented like παῖς. 


(α) The following neuter nouns with -t- in oblique cases 
and plural have a divergent nom. acc. sing. :— 


Tovar- knee (1.6. yovF- ar-) has nom. γόνυ, a -%-stem, which was declined 
throughout in early Greek. 

ὮὯτ- car has in papyri of iii/ and ii/B.c. a nom. acc. és, levelled from 
οὖς by the influence of ὦτα, ὠσίν etc. (Mayser 5). It is not found in NT. 


Two remain of the very ancient declension which had -p 
in nom. acc. sing. and -atos ( = -n-Tos) in gen. :— 


‘ySar- water has nom. ace. ὕδωρ : the whole flexion occurs in NT. 


Φρεᾶτ- (for ppnar-) well, nom. ace. φρέαρ, gen. φρέατος. It is rarely 
contracted (φρῆτος) in the vernacular. 


(6) One -θ- stem survives, opvi0-, which has specialised 
its meaning just as our fowl has. Ὄρνεον or πετεινόν replace it 
in the wider sense bird. It only occurs once, Mt 23°" ὄρνις 
nom.: the parallel passage in Lk 13* has probably (see (1) 
above) the dialectic variant ὄρνιξ. 


In ABLR al. the reading is assimilated to Mt. If WH are right in 
calling dpvié “ Western,” we have a small point which might go towards 
locating this type of text in Asia Minor, with Egypt (on the papyrus 
evidence) as an alternative. More probably ὄρνις is ΜΒ form and ὄρνιξ 
that of Luke, who has again refrained it would seem from altering an 
“incorrect” form of Q. But possibilities are many and evidence 
ambiguous, so that we must be cautious in inferences. 


134 A GRAMMAR OF NEW TESTAMENT GREEK. [8 55-56 


(4) Stems in -ντ-. 


These form a special class, because of the phonetic results 
of the addition of -c- suffixes: there is also the double type 
in nom. sing. from stems in -ovt-, which admits of more than 
one explanation—see Brugmann-Thumb Gr. 257. 


apxovt- (δ) ruler, ὀδοντ- (δ) tooth. ἧμαντ- (ὃ) strap. 


Sing. NV. ἄρχων ὀδούς ἱμάς 
A. ἄρχοντα ὀδόντα ἱμάντα 
G. ἄρχοντος ὀδόντος ἱμάντος 
1). ἄρχοντι ὀδόντι ἱμάντι 

Plur. N. ἄρχοντες ὀδόντες ἱμάντες 
A. ἄρχοντας ὀδόντας ἱμάντας 
G. ἀρχόντων ὀδόντων ἱμάντων 
D. ἄρχουσι(ν) ὀδοῦσι(ν) ἱμᾶσι(ν) 


There are no separate Vocatives. Under Adjectives and 
Participles will be found types of Neuters, and of stems in -εντ-, 
-wvT- and -ovvt-.. The flexion of the noun types ὀδούς and 
ἱμάς is incomplete in NT, but adjectives and participles justify 
their being printed as if complete. 


Latin nouns in -evyr- made nom, in -ης (Lat -éns) : Kphoxns=Crescens 
and Πούδης -- Pudens (nom.), Κλήμεντος (gen.) = Cleméntis (K.BI. i. 491). 
B. Nasal Stems. 


§ 56. In Greek these all end in -v, through the influence 
of a final μ᾽ becoming -v in nom. sing. by phonetic rule: thus 
the very few μ- stems (as χιών, χθών, eis) were assimilated. 


1. Stems with Strong Flexion. 


ποιμεν- (0) shepherd. ἤγεμον- (δ) leader. κυον- (6) dog. 
Sing. N. ποιμήν ἡγεμών κύων 

A. ποιμένα ἡγεμόνα κύνα 

(ἡ. ποιμένος ἡγεμόνος κυνός 

1). ποιμένι ἡγεμόνι κυνί 
Plur. N. ποιμένες ἡγεμόνες κύνες 

A. ποιμένας ἡγεμόνας κύνας 

G. ποιμένων ἡγεμόνων κυνῶν 

1). ποιμέσι(ν) ἡγεμόσι(ν) κυσί(ν) 


There are no Vocatives surviving here (classical in barytone 
words, as δαῖμον, κύον). The voc. ἄφρων is presented: in the 


S$ 56] NOUNS. 135 


old form ἄφρον in Lk 122 KMSUVII etce., 1 Co 15* KL ete. : 
there can be no doubt that the better MSS spell here 
according to Hellenistic use. 


(a) Κύων lost from the earliest times in Greek its middle stem in the 
acc. sing. and nom. pl., where historically it was in place. It is the only 
v- stem surviving in NT which preserves the weak stem, except the 
isolated ἄρνας (ace. pl.) Lk 10%, which comes from the long obsolete nom. 
sing. ξαρήν (found only in two or three ancient inscriptions). In NT 
lamb is ἀμνός (= Lat. agnus—generally, as in classical Greek, in nom. sing.) 
or the derivative ἀρνίον : ἄρνα, ἀρνός etc. are common in LXX, but rare 
in papyri (Mayser 284). Dat. ἄρνασι occurs in literary Κοινή. 

(b) The effects of strong flexion are seen, not only in the lengthened 
vowel of nom. sing., but in dat. pl., where the addition of -σι(ν) to the 
middle stem would have produced -εισι(ν), -ουσι(ν). The weak stem 
leaves its traces here, -ἀσι(ν) (=-y -σι) being assimilated in its vowel to 
the rest of the flexion. f 

(c) About a dozen nouns in NT are declined on these models, and a 
good many adjectives (like ἡγεμών). One or two late uncials in Lk 14!” 
15® spell γείτονας with o. 


2. Stems without Strong Flexion. 


Ἕλλην- (6) Greek. aiwy- (6) age. ὠδιν- (ἡ) throe. 
Sing. N. Ἕλλην αἰών ὠδίν 
A, Ἕλληνα αἰῶνα ὠδινα 
(ὐ. Ἕλληνος αἰῶνος ὠδῖνος 
1). Ἕλληνι αἰῶνι ὠδῖνι 
Plur. N. Ἕλληνες αἰῶνες ὠδῖνες 
A. Ἕλληνας αἰῶνας ὠδῖνας 
G. Ἑλλήνων αἰώνων ὠδίνων 
1). Ἕλλησι(ν) αἰῶσι(ν) ὠδῖσι(ν) 


There are no separate Vocatives here, either in classical 
or in Hellenistic Greek. 


(a) One divergence here must be noted, the levelling of the old nom. 
@dis ef sim. to ὠδῖν, so as to agree with the rest of the flexion. It is 
normal in Hellenistic (cf. Mayser 285, Thackeray 151), and is even found 
in Lucian (δελφίν, K.BI. 1. 415—as voe.), which strikingly shows how the 
old -is had faded out of even literary memory. (W and WS wrongly 
compare κλειδίν from a late writer: it is of course only the vernacular 
form of κλειδίον.) 

(6) Two nouns in -ay- are declined on the above model :—Mey.orav- (ὁ) 
magnate (only plural)—nom. -Gves, dat. -Gou(v). Μελᾶν- (τό) ink (only 
sing.)—gen. μέλανος, dat. μέλανι. It is the neuter of μέλας, μέλαινα, 
μέλαν black (see 8 65 (3) a, p. 160). 


136 A GRAMMAR OF NEW TESTAMENT GREEK. [8 56-57 


(ὁ) Mnv- (6) month has without variant the Attic nom. μήν, not the 
older eis. Cases accented regularly, μῆνα, μηνί, μῆνας. 

(d) Besides those already mentioned, there are fifteen nouns in NT 
declined like αἰών, together with the name Σαλαμῖν- (ἡ), like ὠδίν (Σαλαμῖνι 
dat.—for a heteroclite variant see above, §54 B(d)). The inferior uncials 
and D would add μυλῶνι at Mt 2441. It is worth noting that foreign 
words (ἀρ(ρ)γαβών, λεγιών, κεντυρίων, χιτών) and late formations make up 
the bulk of the list, in which ἀγών, αἰών (whose ancient locative αἰέν ever 
attests original strong flexion), κλύδων (μυλών), χειμών, XLT are the only 
ones that claim classical antiquity. This is due of course to the fact that 
a simple flexion like this was naturally adopted for new words. 

(6) Inferior uncials sometimes spell with o for oblique cases of 
ἀρτέμων, ἀφεδρών, κλύδων. 


C. Liquid Stems. 


§ 57. “Ads, the oniy word with stem in A, has been dealt 
with above (p. 132), so that we are exclusively concerned with 
stems in p, which are very numerous. In this class occur 
the most considerable survivals of vowel-gradation in the 
stem, especially in the old relationship-nouns. 


(1) Nouns with Strong Flexion. 
πατερ- (6) μητερ- (ἡ) Ovyarep-()  ἀνερ. (ὃ) 


Sather. mother. daughter. man. 
Sing. N. πατήρ μήτηρ θυγάτηρ ἀνήρ 
V. πάτερ θύγατερ ἄνερ 

A, πατέρα μητέρα θυγατέρα ἄνδρα 

G. πατρός μητρός θυγατρός ἀνδρός 

1). πατρί μητρί θυγατρί ἀνδρί 

Plur. N.V. πατέρες μητέρες θυγατέρες ἄνδρες 

Α. πατέρας μητέρας θυγατέρας ἄνδρας 

(ἡ. πατέρων μητέρων θυγατέρων ἀνδρῶν 


1). πατράσι(ν) μητράσι(ν) θυγατράσι(ν) ἀνδράσι(ν) 

In the vocative μῆτερ is not given, as it is rather unsafe to assume a 
vocative that does not actually occur. Thus in BGU iii. 846!° (ii/A.D.) 
(= Milligan no. 37) we have μήτηρ as voc. in an illiterate letter. Πατήρ 
and θυγάτηρ as vocatives appear three times each in the Synoptic Gospels 
in the best MSS (four times in Jn): there does not seem adequate reason 
(with WH) to accentuate this πάτηρ (App.? 165), as it is simply a nom. 
used as voce. (cf. υἱός above, ὃ 52). 


Strong flexion originally required the lengthened stem in 
nom. sing., the middle in ace. and locative (here = dat.) sing. 


§ 57] NOUNS. 137 


and nom. plur., and the weak elsewhere. It will be seen 
that when we put the four nouns together these conditions 
appear in one or other of them throughout, except in dat. 
sing., where earlier Greek could use the middle stem (μητέρι, 
avépt). The weak stem ἀνδρ- has a transition sound ὃ 
between τὶ and r (cf. our thunder). The -ρᾶ- in dat. pl. 
represents vocalic 7. 


Γαστερ- (ἡ) (belly), womb, is declined and accented like πατήρ but shows 
only dat. sg. γαστρί (and nom. pl. yaotépes=gluttons in a quotation from 
Epimenides (vi/B.c.) at Tit 112). 

᾽Αστερ- (6) star might historically be placed here, though it has levelled 
away its gen. and dat. sing. into ἀστέρος, ἀστέρι, for its dat. pl. was 
ἀστράσι(ν). But although this occurs in late writers, it cannot be shown 
to survive in the vernacular. In NT, where the whole flexion occurs 
except dat. sing. and pl., it is perhaps significant that ἄστροις appears 
in Lk 2125, where ἀστράσιν would have served equally well. Crénert 
173 quotes ἀστῆρσι and dorépos as MS readings in Geminus, a writer 
of i/A.D. 


(2) Stems with partial Strong Flexion or none. 


pytop- (6)  σωτηρ- (6) σχειρ- (ἡ) paptup- (ὃ) 
orator. saviour. hand. witness, 
Sing. δ΄. ῥήτωρ σωτήρ χείρ μάρτυς 
Α. ῥήτορα σωτῆρα χεῖρα μάρτυρα 
G. ῥήτορος σωτῆρος χειρός μάρτυρος 
72). ῥήτορι σωτῆρι χειρί μάρτυρι 
Plur. N. ῥήτορες σωτῆρες χεῖρες μάρτυρες 
A. ῥήτορας σωτῆρας χεῖρας μάρτυρας 
G. ῥητόρων σωτήρων χειρῶν μαρτύρων 
1). ῥήτορσι(ν) σωτῆρσι(ν) χερσίᾳν) μάρτυσι(ν) 


No Vocatives are found. 


(a) The classical type ῥῆτορ is pointedly set aside in the recurrent 
LXX κύριε παντοκράτωρ ; nor is the old σῶτερ (abnormal in a word with 
-τήρ -ripos) traceable in LXX or NT. 

(b) Papyri and inscriptions guarantee datives like Φιλομήτορσι, Σωτῆρσι, 
and the gen. σωτήρων, which do not occur in NT. 

(c) Two nouns in -ἤρ -épos may be declined after ῥήτωρ mutatis 
mutandis, but with no warrant for a dat. pl. (φράτερσι occurs in Attic). 
"Aep- (6) air has nom. ἀήρ, acc. ἀέρα, gen. ἀέρος. For ἀστήρ see above. 


C (1). 


138 A GRAMMAR OF NEW TESTAMENT GREEK. [§ 57-58 


(d) The real stem of χείρ is yepo-, whence χειρός and χερσί came 
phonetically, the nom. being made up afresh from oblique cases (G. 
Meyer® 414). 

(e) Like paprup- (sing. only) is πῦρ- (τό) fire, with nom. acc. πῦρ, gen. 
πυρός, dat. πυρί. The nom. μάρτυς occurs eight times in NT, with no 
sign of μάρτυρ : for the loss of p see Brugmann Grundr.? τ. 435. 

(f) There are five nouns in NT (one only in nom. sing.) with deel. 
after σωτήρ, and eight or nine after ῥήτωρ. The two in -7p -epos may be 


added, and a compound each of χείρ and μάρτυς. On κατήγωρ see above, 
§ 54 B (ce). 


D. Stems in -o-. 


§ 58. Since original o disappeared in Greek between 
vowels, there are contractions of concurrent vowels in these 
stems everywhere except in nom. (acc. neut.) sing. and dat. pl. 
This class is in Hellenistic practically confined to one type, 
neuters in τος, which in oblique cases show vowel-gradation 


(-eo-). 


(1) Stems in -ος : -εσ-. 


ἐθνεσ (τό) nation. συγγενεσ-- (δ) kinsman. 
Sing. NV. ἔθνος συγγενής 
A. ἔθνος συγγενῆ ( =- ε(σ)α) 
G. ἔθνους (= - ε(σ)ος) συγγενοῦς 
1). ἔθνει (= -εσ-ι) συγγενεῖ 
Plur. N. ἔθνη (= -εσ-α) συγγενεῖς (= -εσ-ες) 
Α. ἔθνη συγγενεῖς 
G. ἐθνῶν (= -εσ -ων) συγγενῶν 
1). ἔθνεσι(ν) (-Ξ- -εσ-σ) συγγενέσι(ν) 


(a) Συγγενής was properly an adjective: its flexion as such is given 
as a model below, p. 162. Note another dat. pl. of this word, συγγενεῦσιν 
Mk 6! B*LA al.® 33 1 ete. (18) ete. al.? (-€ow NACD*W al.), Lk 244 
LWXAA 1 ete. 13 etc. 33 al.!° (-ἐσιν NABCD al.). Cf. 1 Mac 10®°. The 
MSS which give this form in Lk (where all the great uncials have the 
normal form) have evidently been influenced by Mk, whose use of this 
vernacular heteroclisis is characteristic: since the passages are not 
parallel, Luke has not his common motive for using a popular form. 
The plurals of nouns in -7s and -εύς coincided in Hellenistic in nom. and 
acc., and the sing. in dat.; while fluctuating usage in contraction would 
bring together acc. sing. and gen. pl. as well. In this case the plural of 
γονεύς would be the principal force. For other exx. of συγγενεῦσι, also 
συγγενέων, συγγενέας and acc. sing. συγγενέα, see Thackeray 153 n.?, 
Crénert 173. WS 89 cites an Atticist’s ban as evidence for συγγενεῦσιν. 


§ 58] NOUNS. 139 


(ὁ) The ace. pl. mase. and fem. is borrowed from the nom. : other- 
wise -éas contracted to -7s would have been found. 

(c) Gen. pl. without contraction appears in ὀρέων Rev 6'5, and 
χειλέων Heb 1315, apparently with no variants: contracted forms 
occurring are ἐθῶν, ἑλκῶν, κτηνῶν, μελῶν (bis), ἐθνῶν (43 times), ἐτῶν 
(14 times), and in the adjective formation ἀσεβῶν (ter), ἀσθενῶν (bis), 
συγγενῶν (bis). The disparity is very great, but in the two words 
affected the open form seems firmly established : it is always found in 
LXX (ὀρέων some seventy times, χειλέων forty), while ἔτος and σκεῦος 
have -ῶν but τεῖχος usually -έων (gen. pl. not in NT): see Thackeray 
151. For outside evidence see CR xv. 435, Mayser 17, 277, Cronert 172, 
Schweizer 153, Nachmanson 135. It seems clear that the uncontracted 
form was throughout kept mainly for certain words: a priori we should 
assume that these came into use in the Kown especially from districts 
(lonic or other) in which the open forms were normal, but it would be 
hard to prove it. Schweizer himself expresses this as his opinion 
immediately after citing the evidence which shows that -év was normal 
(even τειχῶν) in Pergamum-—-Ivnic territory! Nachmanson’s exx. prove 
the same for Magnesia: we may admit that neither touches ὄρος or 
χεῖλος. The infection did not reach Egypt, where P Tor 131} βλαβέων 
(B.C. 137) is the only early example. See Thackeray 144, 151. 

(d) The addition of -y in ace. sing. masc. or fem. is distinguished from 
the case of -ἄν (above, ὃ 55) by the stronger influence of the Ist deecl., 
combined with the strong tendency to add irrational final ν after long 
vowels. With short finals it was much less common. The -ys nouns in 
ist and 3rd 460]. agreed originally in nom. and dat. sing. (ec and » being 
identical in Attic, though not in Kowy); and the dropping of the gen. 
-s was as easy as the adding of ν. Hence even in iv/B.c. proper names 
of 3rd decl. preferred -nv : Σωσθένην in Ac 1817 is Attic. See K.BI. i. 
512f. But the later extension (ἀσφαλῆν etc.) is less obvious than it 
would seem, for the datives were no longer identical in sound—y and ἡ 
alike were Ὁ, while εἰ was 7 (see ὃ 24). More serious is the difficulty of 
the accent. If the analogy of Decl. I. was still the operative force, we 
should naturally write συγγενήν, like κριτήν ; but it is hardly likely that 
the addition of so fugitive an element—added so recklessly, as we saw 
(§ 47), to all manner of long vowel endings—should have altered the 
circumflex to an acute. We may regard it then as a special case of 
“irrational -v,” encouraged largely by the analogy of other accusatives 
in -nv: cf. the raising of the “freak” ἦν into something like a regular 
flexion form by the fact that it coincided with an existing form in the 
conjugation of εἰμί. It is presumably only a coincidence that in Lesbian 
Aeolic the -ys proper nouns took a flexion modelled on 1st decl. (Thumb 
Dial. 262). The occurrences of this -ῆν acc. in NT may be summarised 
thus : αἰσχροκερδὴν Tit 17 FG, ἀσεβὴν Rom 45 ND*FG, ἀσφαλῆν Heb 619 
ACD*P, αὐθαδὴν Tit 17 FG, povoyerny Jn 316 cu? (including 13-346) 
Heb 1117 D*, ποδήρην Rev 11% A cul, συγγενῆν Rom 161! AB*D*, ὑγιῆν 
Jno NeW 15 a it Ὁ ( 

(9) The Attic ace. sg. ὕγιᾶ occurs in three cursives (incl. 1) at Tit 28, 


140 A GRAMMAR OF NEW TESTAMENT GREEK. [§ 58-59 


(2) Stems in -oo-. 


One noun, once used, survives from this rare declen- 
sion. 

Aidoc- (1) modesty, makes nom. αἰδώς, gen, αἰδοῦς 
(=-oa-os), The type reappears among the adjectives and 
participles, where it accounts for the forms without v in the 
-ἴων comparative, and for some of the perf. partic. act. For 
these see § 65. 

For the similarly declined ἠχώ see below, ὃ 59 (6). 


(9) Stems in -ασ-. 


Two nouns show traces of this type, never common; and 
(as might be expected) levelling has worked even here. But 
see Kown instances in Schweizer Perg. 156. 

Γηρασ- (τό) old age occurs in dat. Lk 156. yjper—so all 
uncials and many cursives. Papyrus instances of yijp-as -ws 
and τᾷ are given in Vocab. sv. Kabel 426 (Christian) has 
γῆρος nom. 


Γήρει is best taken as a simple assimilation to the -os nouns. But it 
μῶν be Lonic, for Herod. has κέρεος κέρει ete., and Homer οὔδεος οὔδει 
from οὖδας. Attic κνέφας has gen. κνέφους. 


Kpeac- (τό) flesh forms κρέα in Rom 1471, 1 Co 818: the 
plural is collective. Herodian says the Kow7 pronounced 
xpéa, as from Homer down (K.BI. 1. 431). Other nouns of this 
class have passed into the -t- class: see ὃ 55 (3) 6. 


δ 59. #. Stems in Semi-vowels. 


(1) Stems in εὐ : and ev: v. 


πόλει- (ἢ) city. σινάπει- (τό) mustard. πηχευ- (0) ell. 
Sing. Δ΄. πόλις σίναπι πῆχυς 
A, πόλιν σίναπι πῆχυν 
G. πόλεως σινάπεως πήχεως ΟΥ̓ πήχεος 
D, πόλει σινάπει πήχει 
Plur. N.A, πόλεις πήχεις 
G. πόλεων πηχῶν (πήχεων) 


1). πόλεσι(ν) πήχεσι(ν) 


§ 59] NOUNS. 141 


Both in form and in accent the flexion of the πόλις 
type presents obvious irregularities, which are however all 
classical. Nouns in -tei (-ovs, except πίστις) form the bulk 
of this class, which includes also one masculine noun, ὄφις 
snake, and a few more feminines. The neuter only occurs in 
one form, and was never more than sporadic among nouns. 
The papyri show the nouns ἄμι, σέσελι, στίμι, κόμμι, πέπερι 
—all foreign, like σίναπι itself. For corresponding forms in 
-eu- we have mostly to refer to the adjectives. The common 
noun πῆχυς is the only one occurring in NT, and there are 
hardly any others in Hellenistic: its flexion is guaranteed 
from other Hellenistic sources. ἄστυ city, the only native 
neuter, was obsolete. 


The influence of the commoner -u- nouns is seen in the form mnyvos 
gen. P Oxy ii. 24215 (a.pD. 77). The LXX gen. sing. is πήχεος (Thackeray 
151), but πήχεως (as Attic) appears in BGU iii. 910 *° (a.p. 71). In the 
plural πηχῶν is always found in papyri, and has the additional recom- 
mendation of being δεινῶς ἀνάττικον for Phrynichus: see Mayser 267. 
It is suggestive that in P Flor ii. 262° (ii/A.D.) πηχεὼων has the e erased, 
The Attic form occurs often in LXX (Thackeray l.c.), and in Jn 218 AW, 
Rev 2117 δὶ, 


(2) Stems in -v-. 


σταχυ- (6) ear of corn. ὀσφυ- (ἡ) loins. δάκρυ- (τό) tear. 

Sing. NV. στάχυς ὀσφύς (δάκρυ) 

A. στάχυν ὀσφύν (δάκρυ) 

G. στάχυος ὀσφύος (δάκρυος) 

1). στἀχυϊ ὀσφύι (δάκρυϊ) 
Plur, N. στάχυες ὀσφύες δάκρυα 

A. στάχυας ὀσφύας δάκρυα 

G. σταχύων ὀσφύων δακρύων 

2). στάχυσι(ν) ὀσφύσι(ν) δάκρυσι(ν) 


Adxpv is an isolated word, for which the heteroclite sing. 
N.A. 8dxpuoy appears in NT. ‘The sing. certainly was 
obsolete in the -v form, and in the plur. it is only the dat. 
(also Attic) that fixes it here. 


K.BI. i. 438 f., 488 follows Herodian in circumflexing sing. N.A. of 
the oxytone words ὁ ἰχθῦς fish, ἡ ὀφρῦς brow and ἡ ὀσφῦς. Brugmann 
Grd? τι. i. 137 does the same. Historically the v is long in these 


142 A GRAMMAR OF NEW TESTAMENT GREEK. [ὃ 59 


oxytona—cf. Skt. bras brow—the dat. pl. being shortened to match the 
barytones. See also Chandler § 620. Ὁ βότρυς bunch of grapes, ἣ ἀχλύς 
mist, ἡ ἰσχύς strength and 4 ὗς sow are the only other words in this 
class. For small traces of the old ace. plur. in -ds for vas see Thackeray 
147; there are no signs of it in NT. 


(3) Stems in -ev-. 
βασιλευ- (6) king. 
Sing. NV. βασιλεύς Plur. N.V.A. βασιλεῖς 


V. βασιλεῦ G. βασιλέων 
A. βασιλέα 2), βασιλεῦσι(ν) 


. βασιλέως 


σ 
42). βασιλεῖ 


So ἃ dozen common nouns in NT. The flexion is like 
that of the later Attic, with no sign of the old ace. pl. in 
-€as. Note that 6 ἁλιεύς fisherman dissimilates ὁ to e before 
the ὁ sound in,plur. ἁλεεῖς, as in LXX (Thackeray 84). 
On heteroclite dat. pl. συγγενεῦσιν from συγγενής see above, 
§ 58 (1). There are no traces of the not uncommon ace. 
sing. βασιλῆ. 

The primitive noun Ζεύς (1H. dytus, gen. diuds, with 
strong flexion) appears in the ace. Ata Ac 14! (Atay DEH LP?) 
gen. Atds. 


(4) Stems in -ov-. ° 
βου- (ὃ) ox. 
Sing. V. βοῦς Plur. N. βόες 
A. βοῦν A, Boas 
G. Bods G. βοῶν 
2). Bot D. βουσίᾳ(ν) 


Ὃ νοῦς mind, ὃ πλοῦς voyaye, ὃ χοῦς dust, have in Hellenistic 
transferred themselves to this class from Decl. II.: νοῦς is the 
only one in NT that has unambiguous forms of this flexion 
(G.D. sing.). See copious Hellenistic citations in WS 84 n., 
also Schmid Attic. iv. 24, 586. 


(5) Stem in -dv-. 

Ἢ vats ship has ace. ναῦν in a passage which seems to 
be a literary reminiscence: ef. Prol. 25 f. The word with 
its medley of irregular forms naturally gave way to πλοῖον in 
the vernacular. 


§ 59-60] NOUNS. 143 


(6) Stem in -οἱ. 
A few nouns, declined in sing. only, followed the norm 


of πειθοι- (ἡ) persuasion, which still is found in a papyrus of 
ii/A.D. (P Oxy iii. 474°"). 


NV. πειθώ (or -ᾧ) G. πειθοῦς ( = -dx0s) 
A, πειθώ (= -όια : accent irreg.) D. πειθοῖ (= -όμ) 


See K.Bl. i. 453f. The flexion concerns us if we regard 
ἤχους in Lk 21% as ἠχοῦς from ἠχώ sound: so WH. See 
above, §§ 54,58 (2). In 1 Co 2* πειθοῖ is an extremely 
probable reading, involving only the dropping of c before 
another c, and the acceptance of the reading of FG omitting 
λόγοις. Πειθοῖ was read by the old Latin and the Sahidic 
and Peshitta: the adj. (e)c0os cannot be proved to have 
existed at all. See ὃ 35 (p. 78). 


DECLENSION OF SEMITIC NAMES. 


§ 60. The Greek Bible presents a very obvious contrast 
to writers like Josephus! in its treatment of Semitic names, 
which are very largely left indeclinable. Thackeray (Gr. 160) 
gives as the general rule for the LXX that 


Names which in the Hebrew end in a consonant remain unaltered 
(Addu, ᾿Αβραάμ, Δαυείδ, Ἰσραήλ, Ἰωσήφ etc.), while those which end in a 
vowel, especially in ΠῚ, are in most cases declined like nouns of the first 
declension, the feminines requiring no addition in the nominative, the 
masculines taking on the termination -ίας and being declined like Νικίας. 
Names ending in other vowels are either Hellenised by the addition of 
s and form a new class of first declension names in -as, -ῆς, -ovs ete. 
(Ἰωνᾶς, Μωυσῆς, Ἰησοῦς etc.), or remain indeclinable (Ἠλειού). 


Since these rules may be transferred to the NT with 


little modification, it will be convenient to follow Thackeray’s 
paragraphs and apply them successively. 


A. Personal Names. 


(1) Indeclinables—The extent to which the use of indeclinable forms 
prevails is well seen in the genealogies of Mt 1 and Lk 3. In the former 
there are 46 names (42 men and 4 women), of which ᾿Ιούδας, Σολομών, 


1The contrast may be well seen in WS 91, where the Graecised proper 
names of Josephus and others are quoted in abundance. I have not thought it 
worth while to repeat them here. 


144 A GRAMMAR OF NEW TESTAMENT GREEK. _ [8 60 


Οὐρίας, “Ofetas, “Elexias, Μανασσῆς, ᾿Ιωσείας, “lexovias and ᾿Ιησοῦς are 
declined: it is curious that ᾿Αβιά (N.A.G.) is not treated like other 
names in ny, but the LXX ΟΛβιού in Kings, ᾿Αβιά in Chr) does the same. 


Of the 75 names in Lk 325% none can be taken with perfect certainty 
out of the indeclinable category, though indecl. Ματταθιού (vv.?5 36) has 
no LXX authority, and Ἰησοῦ and Ἰούδα bis are presumably also from 
nom. in-as. Aevei bis probably is as elsewhere from Aeveis, but there is 
no strong reason for assigning to the Mixed Decl. (see (6) below) the other 
names with vowel endings, A summary may be added of Semitic names 
in NT (incl. ‘AxeASapay, Boavnpyés, yévvea, μαμωνᾶς, Σαβαώθ and χερουβείν) 
as in WH text :— 


Ending in. .]a. e.|>| a €.) 7. | 9 |e. | κι] A.) me] ν- | ρ. | ς- | 7 | ve 6.x ω. 
| Ι 
reat | 
Z| peat ΠΩ cette as [=| | 1... stb ail ccelere hee] oe en =a 69 
2 © | Indecl. w2/4}2/6|4 5 9 | 6 |20/ 20 19/14] 1156} 46 1|-44 
J Ϊ Pet | 

i | 

2 | Dect. ΕΠ 1 ||} 7 oom (real Wee? 3 BS Nesey) Wise aes υὉ = 98 
3 

Ba finder hs | orl eee pee ort lees feta ΝΣ 9/8 211 rH ac 1/= 33 


To these should be added 28 gentilic names and the like derived from 
Semitic words : see (16). 

(2) Declension II.— A few names have been Graecised in this way, viz. 
* (2)”"AyaBos, ᾿Αλφαῖος, * Βαρθολομαῖος, Βαρτίμαιος, *’EAucaios, + Ζακχαῖος, 
* ZeBedaios, Gaddaios, ft "Iderpos, * Ιάκωβός (not the patriarch, nor the person 
named in Mt 115"), Aa¢apos, Λεββαῖος (in 6-text, perhaps a duplicate of 
Aeveis), Μαθθαῖος, Μάλχος, * Σαῦλος (against Σαούλ as name of the king, 
and in a direct citation of Aramaic). Those marked * are names which 
are only indeclinable in LXX; in those with + the NT form is antici- 
pated in LXX, with or without alternative. It should be added that 
Ἰάκουβος occurs once in LXX text (1 Es 948 A). We find Δανιήλου in 
Mt 241° Ὁ, Γαμαλιήλον Ac 223 B. (I have excluded Τιμαῖος above on the 
same principle as Σίμων in (7) below, q.v. : Βαρτίμαιος is counted.) 

(3) Feminines in Decl. I—Here we have “Avva (nom. only), Eva, 
Ἰωάνα (nom.), Μάρθα, Μαρία, Ῥεβέκκα (nom.), Σάρρα, Σουσάννα (nom.), 
also Σαλώμη. As in LXX the gen. and dat. are always -ας -ᾳ : so Μάρθας 
Jn 111, as well as Μαρίας and Σάρρας. The variations in the name 
Mapia are complex. In the gen. Mapias stands “virtually without 
variation” (WH) for all the women so named ; and “ Mary of Clopas is 
always Μαρία (nom.’), as is (acc.!) Paul’s helper (Rom 168),” where, how- 
ever, ND, read Μαριάμ. The mother of Jesus WH always give as 
Μαριάμ (nom., voc., ace., dat.), except in Mt 1°° (BL 911), Lk 219(8*BDR) ; 
but even here NCDWw and ALWw could be quoted if we preferred to 


1WS 92 quotes K.Bl. i. 381, where it is observed that names in Attic 
of foreign origin often keep -@ throughout. Cf. Avddas in (10) below; also 
Τα]μύσθᾳς gen. in BGU iii. 883° (ii/a.D.). 


§ 60] NOUNS. 145 


make the rule absolute. The name appears thus, except in the genitive, 
“usually without important variation.” They make the same rule for 
Mary of Bethany, though here they are content once with 33 as sole 
authority. The perpetual variation of the uncials, even in a continuous 
narrative like Jn 11, is inexplicable. Thus for Μαριάμ nom. in Jn 112° 
there is nothing but 33, in 33 we have BC*EL 33; for it is as ace. the 
range is lessremarkable. Again in Lk 1039 Μαριάμ nom, is in NCLPZW 1 
33, against AB*Dw ; in 45 B and 1 alone support it. Only two or three 
MSS are consistent : B 33 have -p 8/10 while NW have -a -αν 9.10. For 
Mary of Magdala the case is still more complex: see WH App.? 163, 
and add that W has -μ 3/13. Gregory (Ti ili. 116) notes that 8D prefer 
the declined, and BLA the indeclinable form: it will be clear, however, 
that there are great inconsistencies, and a rule seems unattainable. 

Σάπφειρα (dat. -7—see p. 118) belongs to this section if taken directly 
from Aram. NYDY “beautiful.” If it is fem. of σάπῴειρος it is Hebrew 
ultimately, for the name of the gem was naturalised in iv./B.c. Blass 
rightly prefers the former, but thinks the common noun influenced it 
(Gr. Tn. : cf. WS 76). 

(4) Masculine names in ΠῚ (etc.), from the name 7)’. There are 
about a dozen of these in NT, declined like Νικίας, with voc. -a and gen. 
του. This is the Greek rule for -as pure (K.BI. i. § 105. 9). So ‘Avavia; 
Βαραχίου, Ζαχαρίου, ᾿Ησαίου, lepepiov, Ματταθίου, Οὐρίου. The only ques- 
tion arises with Ἤλείας, which in Lk 117 makes gen. λεία NBLW 565** 
(-ov ACD etc.), but in Lk 4558 Ἠλεί -ov without variant. (In both places 
late uncials accent -od or -ov, reading the indeclinable form found in the 
LXX (historical books): that the later LXX books (Mal, Sir, 1 Mac) 
show Ἠλείας confirms the reading of the better MSS in NT). Thackeray 
162 argues the -a genitive “vulgar and late”: this difference of Greek 
culture between the sources which Luke reproduces in these two places 
would be quite in keeping with their style. 

Other masculines of Decl. I. are Μεσσίας, Ἰωάνης, Ἰσκαριώτης 
(Ἰσκαριώθ in Mk and in Lk 616)} Note the dat. Ἰωάνει 4 in WH: it 
will be explained like Μωυσεῖ below. 

(5) The difficulty discussed under this section does not arise in NT, 
᾿Αβιά Mt 17 is necessarily indeclinable, not being gen.: the possible 
accentuation Ἠλειοῦ or -ov was mentioned in (4). 

(6) Miaed Declension.—This is a large type in NT, and need hardly 
be set down (with Thackeray) to Hebrew influence. For although it 
contrasts with the form prevailing in Egypt (Ἰησοῦς Ἰησοῦτος and the 
like), it tallies with a type used in Greek abbreviated names (Anpas -ἢ 
etc.) and in Roman masce. names like ᾿Ακύλας gen. ᾿Ακύλα.2 It is more- 
over identical with the sing. flexion of a mass of MGr nouns—yépovras 
A.G. γέροντα, κλέφτης A.G. κλέφτη, παπᾶς A.G. and Voe. παπᾶ, παπποῦς 


1 Note also Σκαριώθ Mk 3! and Lk 616 D. In Jn Ὁ has’I. ἀπὸ Καρυώτου 
(exc. 67), and in Lk 225 ᾿Ισκαριώδ. 

2 As early as Plato (Phaedrus 274d) we find the Egyptian name Θαμοῦς 
with acc. -odv, gen. -o0; but the crucial dat. does not occur there, so that 
Winer’s citation is not decisive: it is only like νοῦς. 


146 A GRAMMAR OF NEW TESTAMENT GREEK. [8 60 


A.G. παπποῦ. Many of these make plur. in -des, as παπάδες, combining these 
alternative flexions. Boppas is an Attic prototype. The general formula is 
that the bare stem, in a long vowel or diphthong, forms the gen., dat., 
and voc. if any, while -s is added for nom. and -ν for ace. Only a few 
call for note :— 

(a) In -as or -Gs. Ἰούδας (with voc. Ἰούδα) is constant in NT, there 
being no sign of gen. Ἰούδου or of indecl. Ἰουδά. So”Avvas, ‘Apéras, Καιάφας, 
Κηφᾶς, BapvaBas, Ἰωνᾶς, Κλωπᾶς, papwvas, Σατανᾶς, Σκευᾶς, Χουζᾶς, to 
mention only those that have the gen. Σίλας (-αν -a) (=NP ND) is not con- 


tracted from Σιλουανός, which is a case like Σίμων in (7) below (Dalman 
Gramm. 124). 

(b) In -ῆς. Μανασσῆς may as well be counted here, though -ν is not 
attached in ace.: LXX has Μανασσή indecl. for the tribe, which is 
equally possible for Rev 7°. Μωυσῆς has been transferred from the 
Mixed Decl. (normal in LXX) to a new flexion universal in and after 
i/A.D., with very marked Hellenisation. We may conjecture that the 
dat. -ei began the type, as a natural dat. for a nom. in -js, and that the 
circumflexed -εἶ led on to a gen. -έως and even an acc. -έα (Lk 1629 all 
MSS), on the analogy of βασιλεῖ -έως -έα. The flexion thus agrees (as 
WS 94 notes) with that of ’Apns in class. Gk. Μωυσῆν the ordinary ace., 
and Μωυσῇ (Ac 7*4, under LXX influence) are the only survivals of the 
Mixed flexion. See WH App.? 165. Ἰωσῆς makes gen. Ἰωσῆ Mt 2756 
ABC al., Mk 65" ACW, 154° 8*ACW al., #7 CW al.; but Ἰωσῆτος Mk 68 
BDLA 33 13 etc. (Ferrar) 565, 154° 47 same (exe. D at 47). This last 
flexion has abundant analogues in papyri, but is solitary in NT, which 
makes for its genuineness, 

(ὁ) In -εἰς. Aeveis in NT conforms throughout to this type. 

(d) In -ovs. Ἰησοῦς has been assimilated to this class, the LXX dat. 
Ἰησοῖ yielding to Ἰησοῦ (as in the Freer MS of Dt). 

(7) Names in ων. ---ΞΣίμων -wvos is declined fully, but it is rather a case 
of appropriating a‘Greek name of similar sound than adapting a Semitic 
one: Συμεών (indecl.) is the Semitic original. Similarly men named 
Jéshfi could either adapt the name as Ἰησοῦς or appropriate the Greek 
Ἰάσων. The only other name in των that is declined is that of Solomon. 
Thackeray shows that the Hellenising of rind took the following order : 
(a) in orthography (1) Σαλωμών (2) Σαλομών (3) Σολομών, (Ὁ) in flexion (1) 
indeclinable (2) gen. @vros (3) gen. -évos. (See his note as to the phono- 
logical meaning of the successive spellings.) Of (a)(1)and (2) we have 
some traces in NT: thus nom. Σαλωμών Ac 747 NAC, and Σαλομ. in Mt 10 
N* 1, Ac 31! A 6153 δὲ, For (6) (1) there is only Mt 1° Σαλομών ace. τὲ 1, 
Jn 1028 W. For (2) the evidence is considerable. Σολομῶν (so accent) 
-avtos, like Ξενοφῶν -ὥντος and several Egyptian names, is supported 
by late uncials generally, with WA in Mt 1°, C (semel) A Mt 1242, 
CKLW al. Lk 113! bis, NCAKLW ete. Jn 1023; and it must be read in 
Ac throughout—3!! SABCP 1. 33 al. (ςφῶνος DE 104), 613 SA 33 al. (-dvos 
BDEP al.), 747 accent Σολομῶν (WH). The late uncials in accenting the 
nom, perispomenon agree with their preference for -évros, which may 


§ 60] NOUNS. 147 


have been due to LXX influence : it is the only declined form that has 
even begun to appear there. 


B. Place Names. 


(8) Places and peoples.—Thackeray notes that “the Hellenised forms 
largely predominate” here. The table in (1) so far bears this out for the 
NT, in that indeclinable forms outnumber inflected by 43 per cent. in 
place-names, but by 110 per cent. in person-names. 

(9) Place-names in -a feminine.—So Τάζα -av (LXX gen. -ns), Σαμαρία 
-as,! Λύδδα -as (but see (10)), Ἰδουμαία -as, ᾿Αριμαθαία -as, Βηθανία -as etc. 
Xappa -as is found twice in LXX, but xappay indecl. usually, and so in 
NT. As with the four feminine personal names in -a which only occur 
in the nominative (see (3) above), we are sometimes left without con- 
clusive proof of their proper category : see (11)b. A place-name which 
almost becomes a common noun is yéevva, gen. -ns (acc., gen., dat. 
sing.). 

(10) Towns in -a.—Declined as neuter plural alone are Σόδομα (as 
LXX) and Σάρεπτα (acc. only, but LXX gen. -wv). Metaplasmus is 
conspicuous here. Λύδδα is twice acc. Ac 955: 85 RAB 33 (hiat. v.5*), where 
CEw assimilate to the fem. form Avdédas in ν.58 (Λύδδης there in Ew) : 
analogy of other forms proves this form Decl. II. and not indeel. (as Hort). 
Τόμορρα, in LXX and 2 Pet 2° (Σοδόμων καὶ Topdppas), rather strangely 
fails to agree with its constant associate. In Mt 1015 however we have 
Γομόρρων. A non-Semitic NT name which behaves in the same way is 
Λύστρα, acc. -av, dat. -o1s. See Thackeray 167f. and Prol. 48 for sundry 
parallels : add P Grenf ii. 74 (A.D. 302), where we have ἐν Tevrvpy bis, the 
village being elsewhere Τέντυρα neut. pl. Similarly the Zoroastrian 
capital Ragha appears in Tobit with ‘Pdyas (acc.) and ‘Pdyy, against 
“Ῥάγων -οις. WS 93 notes the varying flexion in 1 Mac of ᾿Αδιδά (indecl., 
dat. -ovs), Βαιθσουρά (indecl., fem., neut. pl.), and Ταζάρα (fem. and neut. 
pl.). 

Declined in Ist decl. only are Γάζα -av, Βηθανία (but see (11) 0), yéevva, 
Σαμαρία (see also 14). 

The examples of metaplasmus just given show that in foreign names 
it was rather the rule than the exception. This accounts for the flexion 
of the name Jerusalem when declined. Mt 2? πᾶσα Ἱεροσόλυμα is of 
course fem., as in Tob 144 B, and Mt3° may show the same. ‘These are 
the only places where the word is nom., and the rule may be that Ἵερο- 
σόλυμα in nom. follows the gender of ἡ Ἰερουσαλήμ, and passes into 2nd 
decl. for oblique cases. 

(11) a. Ἱεροσόλυμα and Ἰερουσαλήμ.52 The indeclinable continues the 
LXX tradition, for Ἱεροσόλυμα only appears in Tobit and Maccabees : 
even in these it is only invariable in 2-4 Mac (1 Mac 34, Tob 5%, and not 


1 Σαμαρείας gen. in P Petr ii. p. 14, a 1 apyrus of ili/B.c., relating to a settle- 
ment of Jews in the Fayyfiim, mentioned by Josephus: see Tebtunis Pap. ii. 
p- 401. 

For the breathing see above, § 40. 
VO 1 ΡΑΒῚ Dt τ 


148 A GRAMMAR OF NEW TESTAMENT GREEK. [ὃ 60 


without variants), Usage in NT varies largely, and without very obvious 
motive. It may be presented thus : 


= ἘΠ ia 7 


5 24 | νῷ om 3 Ε Θ = 6 Ξ 

Ξ ΕΞ & 3 <4 2% Ξ τι ὥ a0) ὥ 
Ἰόρούσανη μι 3} eee ὥς. nD τες 17 ΡΠ pte ol ee ll ea 

x PENS A Ed ra | SAL AA es MSRM Soh ial ΚΡῊΣ 8. 
x A ΠῚ τ Pho 9} RRS 
Ὁ 7 gra} ge ἢ 1 | =16 
τ D ΠΣ ἈΠ τ ih tee γ΄ Ξ ιν 
Total 1 26 | 22 | ἢ ΕΞ} 
| 
“Ιεροσόλυμα NV. Ai\Weees, 7 east ΠΣ πεν) ΠΡ ΞΕ eects | Mactan 
Ἰ "ΜΠ ΕΣ 8 ate 
a Gea S|, Bl 5 =11 
AS abe | A MR Pl Ξ-14 
ΠΣ κ᾿ ΠΡ Ay ἢ 9 8 =63 
{ 


The figures are for WH text. The only places where variants have any 
real support are :—'Iepovoadnp Mk 11! A al, Ac 15* NCDE al. (against 
AB 81 vg), Ac 201° NAE 33 al. (against BCD ete.), and some places where 
only the case is concerned. It should be added that the ‘‘ We” passages 
of Ac have Ἴερουσ. 3 times and ‘Iepoo. 4; while Q shows Ἱερουσαλήμ voc. 
(Mt 23°7=Lk 1334), and according to Harnack also in Lk 4° (where how- 
ever Mt has not the name). In writers who use both forms—which does 
not include Mt, since his one instance of the indecl. comes from Q—it 
has been largely assumed that (as Grimm puts it) “a certain sacred 
emphasis . . . resides in the very name” where the indecl.-form is used. 
This may account for a good many passages, but other forces are likely 
to have co-operated. Since Luke uses Ἰερουσαλήμ 48 times in the 
Palestinian narrative (Ev, Ac 1-12) against Ἱεροσόλυμα 9 times, while in 
Ac 13-28 the proportion is markedly reversed (14 : 18), it is clear that 
he keeps the LXX form as congruent with the atmosphere of his story 
till he emerges into the Gentile world, where the Jews’ capital was 
vaguely supposed to have some connexion with the Solymi (neighbours 
of the Lycians) and the name had been fixed by popular etymology. 

(11) b. Indeclinables in -a are AadpavovOa (st v.l.), Kava, Maydada 
(CM boh al.—the true reading is Μαγαδάν), Σινά. Βηθαβαρά!} and ‘Papa 
occur in the dative, and we could write -a@ if we liked; similarly 
ambiguous are Βηθζαθά with its variants (nom.), and Ταββαθά (probably 
nom.?). Τολγοθά is classed as indecl., though Τολγοθάν ace in Mk 15”? 
might be assigned to the nom. Τολγοθά of Jn 1917 and (probably?) 


1 The true reading is Βηθανίᾳ. Burkitt assigns the variant to the influence 
of syrvt. 

2 The syntax is like that of ᾿Ελαιών in Lk 19: see Proleg. 69, 235, and 
Thackeray 28. 


§ 60] NOUNS. 149 


Mt 273%, If it is indeel., the -av and -a forms will have the same relation 
as in the certainly indecl. Βηθσαιδάν and -d: this account is more 
probable. Βηθανία, normally declined, occurs as Βηθανιά indecl. in Mk 11! 
B*, Lk 1929 8*BD* 131, Mk 1112 H. A special difficulty occurs with 
the name of Nazareth: it is written e.g. Na¢apa (7, indecl.) Mt 41° B*Z 
33, Lk 416 NB*= 33 ; Ναζαρέτ Mt 229 NDL, Lk 12° NBL, Ναζαρέθ Mt 2111 
NBCD etc., Ac 1038 NBCDE, and in A and other MSS sometimes Ναζαράθ 
(-dr). See WH App.? 167, Ti on Lk 17°, WH assert that the tangle 
“presents little ambiguity,” and print -ad in Mt 415 Lk 416, -ἐθ later in 
the Gospel story (Mt 2111) and Ac l.c., with -ér elsewhere “certainly or 
probably.” Sanders (p. 21) says that in W Ναζαρέτ occurs four times in 
Lk, who has Na¢apé6 once (41°): this however stands alone in the other 
Evy., except for Mt 211} -ér. Dalman (Gr.?.152) appears to trace the two 
types to ΠΝ) (Nafapa) and ΓΝ) respectively. The exact Greek name 
of this obscure little place, which was never heard of till the Gospel 
story was current, might easily fluctuate in oral and written sources, 

In -n there is ἘΠῚ Βηθφαγή indecl, Ἰόππη, a town in early contact 
with the outside world, naturally took a Greek flexion. 

In - we find Φαραώ and the place-name Ἰεριχώ indecl. 

(12) Place-names in -wv.—Thackeray makes the interesting point that 
these “are declined or indeclinable mainly according to their rank and 
situation on or away from the main routes.” This obviously suits the 
NT names Βαβυλών -@vos, Σιδών -@vas ; nor need Sapova from Σαρών rank 
as a serious exception, though indecl. in LXX. Uninflected are Mayedav 
(Rev 1616 after “Ap—the compound phrase prob. nom.: see above, p. 
148 1.2) and Αἰνών, Σιών. Κεδρων raises a difficulty in Jn 181. In A(S)YA 
123 latt syrr arm we find rod Κεδρών, which would represent the 
indecl. found in LXX. But 8*DW ab sah aeth read τοῦ κέδρου, and all 
the other Greek MSS τῶν κέδρων, with boh Orig Chr. These would 
plausibly figure as independent attempts to regularise the reading of A, 
regarded as Greek ; and so Lightfoot (Babl. Hsswys 174) actually read. 
But it seems better with WH to accept τῶν κέδρων as a Greek popular 
etymology of Kidron: it is needless with them to labour a proof that 
this etymology was correct.' The Hellenised form starts in LXX, 
2 Καὶ 1523, 3 Καὶ 1513, An interesting parallel occurs in Ps 8210, where 
“some inferior MSS” (Lightfoot) have τῶν κισσῶν, making Kishon into 
“ivy brook.” 

(14)? Names of countries or districts—Thackeray shows that these 
were normally expressed by feminine adjectives in agreement with χώρα 
understood. The oldest suffixes were -is (-idos), -(e)ia (-as), and -ἰκή (-7s) 
which are used for places away from Palestine. ᾿Αραβία, Σιδωνία and 
Συρία are NT representatives, with Σαμαρία as an old name of a district 
within Palestine. Φοινίκη is also Semitic in origin, but is of course not 
formed in this way: it is not included in the table above, being 
naturalised very early in Greek language history. About 200 5.0. the 


10. Schrader, Reallexicon ὦ. idg. Altertwmskunde, 926, gives a very 
different account of κέδρος, which originally meant ‘‘ juniper.” 
2 (13) does not concern Semitic words ; nor have we NT instances for (15). 


150 A GRAMMAR OF NEW TESTAMENT GREEK. [8 60-61 


old indeclinables for names of Palestinian districts began to be replaced 
by adj. in -αία and -(e)iris (τιδος): for the latter we may quote the 
appearance in Ptolemaic papyri of adjectives like ξυλῖτις (yn), ἀμπελῖτις 
(γῆ) “land under trees” or “vines,” and Greek names like Τραχωνεῖτις. 
So αἰγιαλῖτις γῆ Ρ Lond 9947 (a.p. 187-8) (=I. p. 134)-- “ἸΔ Πα on the 
border of the lake.” This last has no Semitic representative in NT: the 
-aia form appears in “IToupaia, Γαλιλαία, ᾿Ιουδαία, Ἰδουμαία. “Αβειληνή 
uses another adj. suffix. 

(16) Gentilic names—of tribes, parties and inhabitants of towns or 
districts—are as in LX X formed largely with the adjective suffixes noted 
in (14), all of course native Greek. In -aios we have Γαλιλαῖος, “EBpatos 
(fem. Ἐβραΐς as an adj.), Ἰουδαῖος (whence adj. ᾿Ιουδαϊκός, and adv, -κῶς, 
and the verb Ἰουδαΐζειν and its derivative ᾿Ιουδαϊσμός), Kavavaios, 
Ναζωραῖος, Σαδδουκαῖος, Φαρισαῖος, XahSatos, Χαναναῖος. In -είτης, 
᾿ἘἘλαμείτης, Ἱεροσολυμείτης, ᾿Ισραηλείτης, Λευείτης, Νινευείτης, Σαμαρείτης 
(fem, Σαμαρεῖτις). In -ηνός, Γαδαρηνός, Γερασηνός (Γεργεσηνός), Δαμασ- 
κηνός, Ναζαρηνός, Μαγδαληνός (only fem.): this is hardly represented in 
the LXX. The less used suffixes are (1) -os in Σύρος, (2) -ἰος, Σιδώνιος, 
Τύριος, (3) -ἰσσα (fem.), Συροφοινίκισσα, (4) consonant noun, “Apay, τι. pl. 
"ApaBes (D* “Apajor). 

The variation between Ναζωραῖος and Ναζαρηνός has been the basis of 
much theorising : it must be left to the Semitist to find out loose stones 
in these structures.1 The former is used exclusively in Mt, Jn and Ac, 
the latter in Mk. Lk has both, in 1837 and in 4%4 241% Luke pre- 
sumably took -yvés over from Mk in 4%, and from the sources of his 
Resurrection story in 2415. Mark’s form is obviously more closely 
related to Ναΐζαρά (-ér, -€0): Ναζωραῖος seems to be coloured by some 
popular etymology, or to represent some other word. Dalman (Gr.? 178) 
makes Ναζωραῖος “reproduce “133, from the by-form ns), synonymous 


with 732,” from which “"182 would be expected.” 


DECLENSION OF NON-SEMITIC NAMES. 


§ 61. There is not much to remark in the flexion of 
Greek names, or even of Latin or other foreign names outside 
the Semitic. They may be briefly classified according to 
their declension. 

(a) Decl. 1. and II.—Feminines in -a and -n are Ἑὐοδία (so certainly 
from the context in Ph 45: 3 (adrais)), Bepvixn, XAon and such Latin names 
as Ἰουλία, KAavdia, with place-names ᾿Απολλωνία, ’Axaia etc., Ἰταλία, 
Σπανία, Λιβύη, Μιτυλήνη, Ῥώμη and various others, derived often from 
native names in sundry languages. In the -d class (gen. -ns) we should 


[}See now The Beginnings of Christianity, 1. i. 426ff., Appendix B, 
Nazarene and Nazareth, by G. F. Moore.—Ep.] 


§ 61] NOUNS. 151 


put Nuyda (see Proleg. 48), Σάπφειρα (§ 60 (8): ultimately, and perhap: 
immediately, Semitic), Τρύφαινα, Τρυφῶσα, and Latin names like Πρίσκα 
and Πρίσκιλλα (only nom. acc.), Δρούσιλλα (dat. -7) 31 ᾿Αντιόχεια, Βέροια. 
In some of these the complete equivalence of ὁ and εἰ in Hellenistic 
makes the spelling, and therefore the accent in nom. and acc., indeter- 
minate. There is no adequate reason to alter the classical spelling where 
we have evidence of it: see Proleg. 46f. and § 35. The plural names 
᾿Αθῆναι, Keyxpeat, Κολοσσαί, Συράκουσαι are classed here. Neuter plurals 
in -a are Θυάτειρα, Avotpa, Mippa, Πάταρα. But metaplasmus is here 
strongly felt : see above, § 54. 

Masculine person- or place-names in Decl. I. include (a) Greek names, 
normally declined, such as Aivéas, ᾿Ανδρέας, ‘Adpias, “Αἰδης, Λυσανίας, 
Ἡρῴδης, Σωσθένης, Ἑὐφράτης (old Persian Ufrdtu, in Greek since 
Herodotus), ᾿Ασιάρχης (see § 54) and two in -irns; (b) Greek and Latin 
names with gen. -a or -a, which might be classed with the Mixed 
Declension. See § 60 (6) above. 

Second Declension names are naturally numerous. Masculine 
personal names in -os call for no comment. Many of them of course are 
Latin, coming from Decl. II. names. Place-names of towns or islands in 
-os are feminine, as ἼΑσσος, Ἔφεσος, but plurals masculine, Φίλιπποι and 
Ποτίολοι (both indeterminate in NT). Among names of countries we 
have ὁ Πόντος (gender indeterminate in NT), but ἡ Αἴγυπτος (Ac 7}! 
and probably °°—Blass’s “wrong reading” in the former only disagrees 
with his own conjectural emendation). Adjectives are numerous, as 
“Αδραμυντηνός, ᾿Ασιανός, Bepouaios, “EAAnvixds. The Mixed Declension 
rather than the obsolete “ Attic” is responsible for ᾿Απολλώς ace. -ὦ or τών 
gen. - (Blass -@s -ῶν -6), and for ἡ Κῶς acc. Κῶ : see § 52 C(c). Neuter 
place-names are Ἰκόνιον, Ἰλλυρικόν etc. Πέργαμον acc. may be from 
either τὸς fem. or -ον neut.: “7 I. in Xenophon, Pausanias and Dion 
Cassius, but τὸ II. in Strabo and Polybius and most other writers and in 
the inscriptions ” (Swete on Rev 213). 

(b) Decl. III.—Normal consonant nouns, Greek or Latin, such as 
Φηλιξ -ixos, Φοῖνιξ -ika, Αἰθίοψ -οπος, Κρής -τός, Ἑλληνίς -idos, Καῖσαρ 
-apos, Μνάσων -wvos, Γαλλίων -ωνος, Ἕλλην -os, need only be named. 
On Σαλαμίς (metaplasmus) see above, ὃ 54. With strong flexion we have 
Μακεδών, Ἰάσων and Φιλήμων -ovos, Φλέγων -οντος, Νικάνωρ -opos. Stems 
in semi-vowels are Νηρεύς, ᾿Αντιοχεύς and other gentilic adjectives ; com- 
pounds of πόλις, declined like the noun;? Svpris and Σάρδεις (pl.); 
Στάχυς (acc. -υν) and Ζεύς ace. Δία gen. Διός. It should be noted that in 
Ac 16"! Νέαν Πόλιν NABD? (against CD*w) and Col 413 Ἱερᾷ Πόλει 
(where MSS are indeterminate) the writing divisim agrees with earlier 
Greek rules: cf. ”Apesos Πάγος Ac 1719: 33 (whence regularly ’Apeomayeirns 
2b.3*), See on this rule further § 106 below. e 

A special case under this heading is the name of the Mount of Olives, 


1 But note gen. Lexdvdas P Oxy ii. 2949 (A.D. 29). 
5 Ἱερὰ πόλις and Νέα πόλις are best written divisim: see § 106. For 
Πρόπολις see Proleg. 228. 


152 A GRAMMAR OF NEW TESTAMENT GREEK. [8 61-62 


on which it will be enough to refer to the discussion in Proleg. 69, 23¢ 
(Einl. 104f.). There I have tried to show that ᾿Ελαιών, a common noun 
=oliveyard,! occurring very frequently in the papyri, is beginning to be 
used as a proper name to be a short substitute for τὸ ὄρος τῶν ἐλαιῶν : we 
are to print πρὸς τὸ ὄρος τὸ καλούμενον ᾿Ελαιών (nom.) or ᾿Ελαιῶνα with 
W in Lk 1939 and (εἰς τὸ κτλ.) 2157, and retain ᾿Ελαιῶνος with all MSS 
in Ae 113, the καλούμενον being an indication, common to Luke and 
Josephus, that the adaptation had not yet thoroughly established itself. 
The Gender of Proper Names shows few irregularities. Towns are 
fem., except when the suffix determines otherwise ; streams follow 
ποταμός" and hills ὄρος in gender, but instances are few. Among 
personal names naturally the gender determines itself; but there is the 
curious Q’r? perpetuwm of ἡ Baad Rom 114, and always in LXX in the 
later books (Chron, Prophets and Tob): in Gen—4 K only thrice (4 K 
1% 16, 918) except asa variant in A only. (In Proleg. 59 (=EHinl, 88) I 
have unaccountably given it as occurring only three times in LXX.) 
The Q’r? is actually written in 3 K 1819. 35 of προφῆται τῆς αἰσχύνης, and 
in the marginal gloss in Q at Jer 1115 ry αιἰσχυνὴ θυσιαστήηριαι The 
explanation, due to Dillmann, ‘‘has superseded all others” (SH on 
Rom 114). Χερουβείν is neut. pl. in Heb 9°, as in Philo and generally in 
LXX, presumably following the association with ζῷα. LXX has χερούβ 
sing. masc. four times, once neut.; χερουβίμ (when treated as pl.) mase. 
four times, neuter twenty. Josephus Ant. viii. 37°* has ras χερουβεῖς. (In 
Ant. iii. 5'87 WS wrongly infers oi x. : αὐτούς there refers to πρόστυποι.) 


MISCELLANEOUS FOREIGN WoRDS AND IRREGULAR FLEXIONS. 


§ 62. In addition to the proper names already catalogued, there are 
a few Semitic words in NT, written in Greek characters, which may be 
collected here, though some of them are not nouns. 


(a) Indeclinable nouns, or vocatives.— Αββά (SBN Aramaic of 38 in stat. 
emphat.) is exactly translated by 6 πατήρ, used in address: see Proleg. 
233 ; Ῥαββεί (D1 Aram.) διδάσκαλε Jn 198 and Ῥαββουνεί (23 Aram. 
“mein Gebieter,” Dalman Gramm.? 176); Ῥακά, an Aramaic term of 
contempt (δ ΡΥ, 7b. 173)®—all these appear only in address, and do not 


develop into regular nouns. The question has been raised whether Mapé 
in Mt 52 is to be read as Greek—so=“ Fool!” in the ethical sense (RV 


text)—or as the Hebrew 77): cf. Num 2010, where LXX οἱ ἀπειθεῖς. 
Field (Notes in loc.) observes that no other pure Hebrew word appears in 


1 How ferfile this formation was in the vernacular may be seen in the list 
collected s.v. ᾿Ελαιών in Vocabulary. 

2 Thus τὸν Σιλωάμ ter: contr. Josephus Wars ν. 4°, 127, ἡ Σ., se. πηγή. 

2 Dalman suggests that the word took the form ῥακά instead of ῥηκά because 
of similarity of sound to ῥάκος ‘‘rascal.” It must be admitted that this is a 
sufficiently rare meaning of ῥάκος (ib. 173 n.”). 


§ 62] NOUNS. 153 


NT except through the medium of the LXX; while Zahn (Matthiius 
p. 225n.) shows that Jewish Midrash writers took min ἴῃ γοῦ. as, ἃ 
Greek word. This seems decisive, added to the obvious consideration 
that Mt’s Greek readers would naturally have been warned if the author 
had meant them to take the familiar word as Hebrew. Πάσχα (Aram. 


NNDB), μάννα (NID), and σίκερα (Aram. N71, according to Dalman! Gr.! 
126 ἢ. 3) are indeclinables. So also the plural XepovBeiv (Heb. DIS, 
with Aram. pl. j.”): see above,§ 61. This last has been included among 
the proper names. So has ᾿Ακελδαμάχ, in Ac 1! tr. χωρίον αἵματος 
(N14 ΡΠ), where Dalman 202 says the final -y is sign of the indeclin- 
able.2 The NT has no trace of declined forms πάσχων, μάννας or 
-ns, aikepos OY -aros, of or ai Χερουβεῖς, which are found in sundry 
writers (WS 91f.); nor again of the Hebrew form φασεκ for πάσχα 
(LXX.) 

(Ὁ) Nouns with Greek suffixes and flecion.—Ap(p)aBav, -dvos (see 8 41) 
was borrowed in iv/B.c. from Pheenician (Heb. 37). Téevva (origin- 
ally yai., as in Jos 1816 B)=Heb, 013773, Aram. 0373, with final ἢ 
dropped (Dalman ? 183): it has been included among place-names above, 
as also Σατανᾶς among the person-names (Heb. }0D, Aram. N]DD). So 
too μαμωνᾶς (=fipxo “ deposit,” according to Dalman? 170 n.), and σαβαώθ 
(pl. of SIS “army”). KopSavas “treasury ” (Mt 27°) is inflected (Aram. 
NIDP), while κορβάν, in Mk 7™ tr. δῶρον, is in Dalman’s view ((r.? 
174) a Hebrew word (131). In Decl. II. we have βάτος (Π3), κόρος (75) 
and odrov (OND), all Hebrew measures, but the form of the last clearly 
depends on Aram. NOND, So with the thoroughly Hellenised σάββατα 
—for decl. see § 54 c—which Dalman (? 160) thinks was spelt with τ for 
8 in virtue of its Greek ending. Συκάμινος is supposed to have been 
borrowed from Heb. TP’, at least as early as Theophrastus (iv/B.c.) : 
popular etymology doubtless affected it. That in OT it is exclusively 
plural suggests that an Aram. pop started the form: the sing, NOP'Y 
is cited in BDB. 

(ὦ) Semtic quotations.—Some fragments of original Semitic language 
appear in NT, simply transferred as spoken. From the zpsissima verba 
of Jesus we have ᾿Αββά (glossed 6 Πατήρ), ῥακά (see (a) above), ἐφῴφαθά he 
opened (NINBX—Dalman? 278 n.), Ταλιθὰ κούμ (2 ΠΡΟ, in which the 
"τ΄ had become silent—see Dalman? 150, 321), and PAGE BNaL hapa 
σαβαχθανεί. In this last (Mk 1534 WH) there has been slight Hebraisa- 
tion, which D carries further by substituting ζαφθανεί for the verb. See 
Dalman? 156, 221, 365: he makes the Aramaic original to have been 
*NPAY ND: ‘PN DN, The bearing of Mark’s authentic record on the 


question of our Lord’s ordinary language is obvious: had it been words 


1 Apparently dropped in ed.?—by oversight ? 
2 See however WS 68 n., quoting Kautzsch. 


154 A GRAMMAR OF NEW TESTAMENT GREEK. [δ 62 


of a sacred text that rose to His lips, we should have had Hebrew—the 
Aramaic attests the speech in which He most naturally expressed Him- 
self when there was no question as to making others understand. In 
this category of Dominica verba we should include ἀμήν truth! It 
established itself in the Christian vocabulary because of its characteristic 
use by Jesus: Jn shows (cf. also Mt 5%7) that He was wont to double it 
for emphasis. Dalman? 183 gives JN as Aramaic, but notes (243) that 
Rabbinic literature has no parallel to such a phrase as ἀμὴν λέγω ὑμῖν. 
‘Qoavvd save! (δ 3 ΜΠ, Dalman? 249) comes from the Gospel story, but 
was well established among the Jews. Another Jewish liturgical term 
taken over was ἁλληλουιά praise ye Jah! (= 73921—Dalman? 191 n.?). 
Mapavaéa is of special interest, as found only in a letter addressed to 
Greeks (1 Co 1653; cf. Didache 10°), Dalman (Gr.2 152 n,8) makes it 
NM NIV our Lord, come! (ἔρχου Κύριε in Rev 22°), See Findlay in loc. 
A password in a foreign language, which embodied the Christian hope 
so as to be unintelligible to the uninitiated, is a very natural and 
suggestive touch in the picture of the primitive Church. Dalman’s note 
seems to dispose of an alleged grammatical difficulty. 

Variants in MSS where these Semitic words are reported belong to a 
Semitic grammar: I have merely reproduced Dalman, with the ordinary 
pointing in place of the supralinear vowels. 

(ὦ) Ina category by itself comes Rev 1* ἀπὸ ὁ dy καὶ ὁ Av καὶ ὁ ἐρχό- 
μενος. It is deliberately left in nom. after ἀπό “in order to preserve the 
immutability and absoluteness of the divine name from declension” 
(Moffatt). The writer aimed at focusing in a phrase the LXX and 
Targums view of Ex 3": cf. also Heb 138. A further tour de force 
makes “the He was” serve as correlative to the present ὁ dv, there being 
no participle to express the continuous past. A Greek might have said 
ἀπ᾽ ἐκείνου ὃς ἔστι καὶ ἦν καὶ ἔσται : cf. the well-known line— 


Ζεὺς ἦν, Ζεὺς ἔστιν, Ζεὺς ἔσσεται" ὦ μέγαλε Zed.) 


Winer’s parallels from Greek philosophical writing—pera τοῦ ἕν, 
χωρὶς τοῦ ἕν, τοῦ μηθέν (Aristotle Politics viii. 3!°—Congreve? p. 347) etc, 
(WM 79)—illustrate the idea underlying the indeclinable, but the 
presence of the article regularises the grammar. Erasmus (supported 
subsequently by two cursives) performed the same service to the text by 
conjecturing ἀπὸ τοῦ ὁ ὦν. For the solecism deliberately conveying 
dogma we may compare Charles Wesley’s couplet— 


The Father, Son and Holy Ghost 
Is ready, with the shining host. 


‘Cf. Preisigke Summelbuch 1540, a tombstone of a.p. 408, beginning θεὸς ὁ 
παντοκράτωρ, ὁ ὧν προὼν καὶ μέλλων, which is a Christian translation of Rev 14 
into idiomatic Greek. Cf. also Syll. 757 (reign of Augustus), where Αἰών as a 
divinity is described, ὁποῖος ἔστι καὶ ἣν καὶ ἔσται, and again ἀρχὴν μεσότητα 
τέλος οὐκ ἔχων. : 


Ot 


8. 63-64] ADJECTIVES. 1 


LATIN Nouns. 


8 63. Proper names are very numerous. We have praenomina, such 
as Gaius,! Marcus, Lucius, Titus, Tiberius, Publius ; nomina, as Quirinius, 
Sergius, Titius, Iulius, Porcius, Cornelius; cognomina, as Paulus, Caesar 
etc. ; agnomina, as Augustus ; women’s names, as Prisca, Iunia, Drusilla, 
Iulia. Words of the 1st and 2nd declensions in Latin fall naturally into 
the same class in Greek : masculines in -a make nom. -as, and gen. -a, as 
᾿Αγρίππας -a (ultimately Greek), ᾿Ακύλας -a (in papyri—gen. does net occur 
in NT), “Adpias (dat. -a). Names of the 3rd decl., as Caesar, Feliz, Gallio, 
are generally simple; we need only note that -éns is transliterated -ns, in 
agreement with the pronunciation (n before s being evanescent), so that 
Κρήσκης and Πούδης nom., Κλήμεντος gen. belong to the same flexion. 
Place-names are also simple, but the Latin -@ is not maintained as it is 
in person-names. Hence Roma Ῥώμη : of course it is -a when pure, as 
Kawapia (so write, preserving the accent of Caesaréa), Ἰταλία. Two Latin 
place-names form a link with the common nouns—’Arziov Φόρον = Appi 
Forum and Τρεῖς TaBépvac= Tres Tabernae, which are of the same kind as 
Καλοὶ Améves, Market Drayton, Sevenoaks, Moses Gate. 

Common nouns borrowed from Latin include :—Decl. I. κολωνία, μεμ- 
βράνη," κουστωδία (κοσ[τ]ωδε[ίᾳ] dat., P Oxy ii. 294°), and the masculines 
κοδράντης (possibly starting from acc. xodpavtny=quadrantem ?), ξέστης 
(if this really is a Latin word) ; also the curiously transformed φελόνης, 
which NT and MGr (edo) show for davoAns=paenula, which is found 
with the other in papyri. For the metathesis see Brngmann Καὶ ΚΟ 249. 
The Keltic (Gaulish) réda passed into Greek (pedav gen. pl. in Rev 1813) 
from Latin. Decl. II. includes κῆνσος, τίτλος, μίλιον (a new formation 
from the plural milia (passuum)), πραιτώριον and others. In Deel. III. 
are words in -wy, of which only λεγιών -ὥνος need be named: in papyri 
λεγεών is also found, but λεγιών predominates.4 Σπεκουλάτωρ makes its 
acc, σπεκουλάτορα, according to the normal flexion of nouns in -ωρ. 


ADJECTIVES. 


§ 64. Adjective flexions need not generally be presented 
in full, as they simply combine types which have already 
appeared among the nouns. We have only to classify the 
types and note some irregularities. 


* Not Caius, which is a pure blunder, due to the misunderstanding of the 
archaic abbreviation C, 

* No early ex. of the sing. is quoted. 

3A vulgar Latin xeata=sexta might support a kind of metathesis: see 
Brugmann Grd.? 1. 871. But the difficulties are great—see also Brugmann- 
Thumb Gr. 159, where the connexion with ἕξ is queried. 

4 See CR xv. 33, 4834: many more could be added—see Vocab. s.v. 


156 A GRAMMAR OF NEW TESTAMENT GREEK. [8 64 


I. SEconD AND First DEcLENsIONS. (Three Terminations.) 


Thus καλός noble, ἅγιος holy, forming the commonest 
type: 


N. καλός καλή καλόν ἅγιος ἁγία ἅγιον 
V. καλέ ete. ἅγιε ete. 


like φίλος... φωνή... τέκνον like φίλος... ἡμέρα... τέκνον 


Contracted adjectives, from stems in which ¢ or o precedes 
the termination, may be declined by combining the flexion of 
διπλοῦς (§ 52 B) with that of γῆ or μνᾶ (ὃ 51), according 
as the last syllable of the contracted word is impure or pure: 
thus χαλκοῦς -ἢ -οῦν brazen, χρυσοῦς -ἢ -odv golden, ἁπλοῦς -ἢ 
-οῦν single, διπλοῦς -ἢ -οῦν double, τετραπλοῦς -ἢ -οῦν fourfold, 
but ἀργυροῦς -ἃ -οῦν silvern, σιδηροῦς -ἃ -οῦν iron: πορφυροῦς 
(only in neut.) would follow the same flexion. For irregulari- 
ties in this respect, and for the appearance of uncontracted 
forms, see §§ 51,52 £(b). The contracted forms are perispome- 
non throughout. Νέος and στερεός, ὑπήκοος and ὄγδοος do 
not contract (the last-named sometimes τοὺς in papyri). On 
ἵλεως (only nom. sing. mase.), the solitary survival of the 
“ Attic Declension,” see above, ἢ 52 Οὐ (Ὁ). “Avirews merciless 
appears in nom. sg. fem. in Jas 215 w, but ἀνέλεος (NABC 
etc.) is the true Hellenistic form. For a discussion of στείρᾳ 
(Lk 156) see below, II. (a). 


II. SEconpD Decienston. (Two Terminations.) 


An adjective flexion in -o¢ -ov, like those given above 
with the separate feminine dropped, belongs regularly to 
compound adjectives, eg. ἡ γυνὴ ἡ ἄγαμος (1 Co 7*) the 
unmarried woman. Exceptions were allowed especially where 
the fem. would be -a (pure), not -7. But there are also a 
few simplicia, which Brugmann (Grundriss* 11. 11. 105) explains 
as being originally nouns. We should not press this in the 
case of adjectives falling into this class in the later language. 
It should be noticed that the tendency of the language set 
ultimately towards eliminating the class: in MGr “all 
adjectives have a separate form for masc. fem. and neut.” 
(Thumb Handb. 67). We must pause for comment on 


§ 64] ADJECTIVES. 157 


some compounds which have taken separate fem. form, and 
on simplicia falling into this class IL, noting also variations 
from earlier Greek. 


(a) Simplicia with two terminations. 

Adjectives in -cos admit of both flexions, even when compound (Blass). 
Kalker p. 239 remarks that Polybius uses παραπλήσιος fem. simply to 
avoid hiatus: this will show that the choice was often very free. In 
this class are— 

Αἰώνιος -ov very often (52 times, of which 43 with ζωή), as usually 
in Attic. But. αἰωνίαν occurs twice, 2 Th 210 (exe. FG), He 9!*; also 
Mk 10% B, Ac 1348 B, 1 Jn 225 B, 2 P 11! Οἱ, 49, 

Kéopuos has dat. fem, κοσμίῳ 1 Ti 29 N*A (al. κοσμίως) : Att. -a. 

Μάταιος has fem. -os in Tit 39 and Jas 12°, but -a in 1 Co 1517 and 
1 P 138; our classical texts show similar fluctuation even within the 
same book 

Νηφάλιος has -ovs acc. fem. pl. 1 Ti 34%. So in Plutarch: normally 
-OS -a -OV, 

Ὅσιος apparently has acc. fem. pl. -ovs in 1 Ti 2%, except in some 
cursives (incl. 33 and 1). Here Winer admitted the possibility of 
Fritzsche’s construction (ὁσίους with emaipovras): against this W. F. 
Moulton referred to Ellicott im loc. The fem. has no parallel here, 
not even in LXX, but an isolated slip, affected by the analogy of other 
adj. in -ἰος fem., is not strange. 

Οὐράνιος (Att. fem. -ia) makes gen. fem, -ov in Lk 215 (but οὐρανοῦ in 
B*D*), dat. τῳ Ac 2619. 

Σωτήριος, as in earlier Greek, has nom. fem. -os (Tit 911), 

On the other hand— 

Βέβαιος has fem. -a alone, Attic showing -os also. But 

᾿Επάρχειος dat. -a Ac 25! B*C is of course the noun ἐπαρχία, ἐπαρχείῳ 
S*A being the adj. in its regular form (ἡ ἐ. sc. ἐξουσία). 

Ὅμοιος has fem. -a except in Rev 4°, where however the agreement of 
ὅμοιος With fps is only an instance of the writer’s normal defiance of 
concord : he has θάλασσα ὑαλίνη ὁμοία κρυστάλλῳ in v.®. Similarly in 919 
ὅμοιοι in two cursives needs no explanation. 

Other two term, adj. in -wos are ἀίδιος, αἰφνίδιος, ἐπίγειος, ἐπουράνιος, 
παράλιος, all in accord with earlier Greek. 

Of adjectives not in -cos we find— 

"Epnpos fem. always -os (Attic also -7), but as a noun (sc. γῆ) except in 
Ac 120 826 Gal 427 

"Eroos fluctuates as in Attic—fem. pl. -o. in Mt 25! (exe, A), else- 
where -os -ἢ -ov. 

Σεμνός has ace. fem. in -ovs 1 Ti 311 A—a mere casual slip. 

Sretpos -ov (so in classical Greek) is a peculiar case, It seems best to 
regard the fem. as στείρα in NT, since στεῖρα would have made dat. στείρῃ 
(see p. 118). The assumption that the NT form is simply a new fem. 
attached to the old adj. of two terminations, and not the independent 
fem. noun στεῖρα (Hom, cf. Skt. starz), cuts out the only exception to the 


158 A GRAMMAR OF NEW TESTAMENT GREEK. [8 64-65 


rule that -pa makes -ns -n. Cf. Gothic staird fem,, which would answer 
to *orépa in Greek. 

Other simplicia with fem. in τος are ἁμαρτωλύς, βέβηλος, φλύαρος, and 
those in -ἰμὸος (φρόνιμος, ὠφέλιμος), all in accord with the sole or pre- 
dominant usage in earlier Greek. 


(Ὁ) Compounds with three terminations are— 


᾿Αργός (-Ξ ἀ-Ε-εργός), nom. fem, dpyn Ja 27°, dpyai in 1 Ti 535, as well 
as in the line of Epimenides Tit 11%, which establishes it as older Greek 
if the reading has not been assimilated. The ban of Phrynichus rests on 
fem. dpyn, which Rutherford (NP 185) thinks may be genuine in Xeno- 
phon : Lobeck, Phr. 104 f. gives plentiful exx. of dpyn in later literature. 

Αὐτόματος fem, -ἡ in Mk 438, Ac 12!° (“not unclassical,” says Blass). 

Καθημερινός, like its classical predecessor καθημέριος, has three termi- 
nations (Ac 6!). 

Παραθαλάσσιος Mt 4: -ίαν (-ἰον 1), παρὰ θάλασσαν N*), but ἡ παράλιος 
Lk 61%, 


III. Torrp AND First Deciensions. (Three Terminations.) 


§ 65. Third decl. stems form their fem. with the suffix 
-(t)i@ : -i-, which in Greek becomes -sa@ gen. -ης. It will be 
convenient to include participles here. 

(1) Stems in -ντ- (cf. ὃ 55 (4) above for the nouns). 


(a) Stems in -αντ-. 


mavt- all. Sing. Plur. 
Sing. NV. πᾶς πᾶσα πᾶν πάντες πᾶσαι πάντα 
Α. πάντα πᾶσαν πᾶν πάντας πάσας πάντα 
G. παντός πάσης παντός πάντων πασῶν πάντων 
D. πανί πάσῃ παντί πᾶσι(ν) πάσαις macy) 


So ἅπας, and all participles in -as. 
(0) Stems in -εντ- 


So participles like τεθέντ- (pres.), πεισθέντ- (wk. aor.). 

Nom. sing. -eis -εἶσα -év; gen. -έντος -eions; dat. pl. 
-elai(v), -είσαις. 

Adjectives in -εἰς -eooa -εν are not found in NT and 
seem to be obsolete. 


(c) Stems in -ovt-. 


So éxovt- willing, dxovt- unwilling (orig. participles) and 


§ 65] ADJECTIVES. 159 


participles in -ovt-, as apyovt- ruling (pres.), ἐδόντ- (strong 
aor.) having seen. 


Nom. sing. -wv -ovoa -ov; yen. -ovtos -ovens. 
-wv -ovaa -ὀν; -OVTOS. 
Dat. pl. -ουσι(ν) -ούσαις. 
-οὔσι(ν). 


One or two formations (from non-thematic verbs—see 
§ 86) make nom. sing. masc. in -ovs, as διδούς pres., δούς 
aor., from δίδωμι give (stems διδόντ-, δόντ-). 


(d) Stems in -ουντ- ( = -εοντ- or -oovT-). 

From Contract Verbs in -éw- and -ow: thus φιλουντ- 
(φιλέω love), δηλουντ- (δηλόω make clear). 

Nom. sing. -ovoa -οῦν; gen. -ovvtos -ovons; dat. pl. 
-οὔσι(ν) -ούσαις. 

(6) Stems in -ωντ- (= -αοντ- or -ηοντ-). 

From Contract Verbs in -ἄω and -jw : thus τέμωντ- (τιμάω 
honour), ζωντ- (ζήω live—entered as Saw in lexica). 

Nom. sing. -@v -@ca -ῶν; gen. -ὥντος -wons; dat. pl. 
-ὥσι(ν) -woass. . 


(7) Stems in -υντ-.- Only two-forms survive in NT, 
from participles like δεικνύς, showing. 

Nom. sing. -vs -ῦσα -bv; gen. -ύντος -vons; dat. pl. 
-ῦσι(ν) -voats. 


(2) Participle stems in -uos- : -us- (-uot-). On the primi- 
tive stem-mixture here see Brugmann (rd? τι. i. 563 ff. In 
addition to the normal form used in the perfect partic. act., 
there is a type of which ἑστώς standing (from ἵστημι) is the 
only NT survival in which contraction has taken place, and 
a new fem. introduced from the present participle. So 
decline εἰδώς knowing and ἑστώς. 

Sing. NV. εἰδώς eiovia εἰδός ἑστώς ἑστῶσα ἕστός 
G. εἰδότος εἰδυίης εἰδότος ἑστῶτος ἑστώσης ἑστῶτος 
Pl. 2). εἰδόσι(ν) εἰδυίαις εἰδόσι(ν) ἑστῶσι(ν) ἐστώσαις ἑστῶσι(ν) 
On εἰδυίης see above, § 49 (2). ‘Eords shortens its proper ὦ (from -aFo-) 


by analogy. In Rev 5°S and some cursives have a neuter ἑστηκώς, which 
if genuine will be due to the writer’s peculiar concord: ὦ and o were 


160 A GRAMMAR OF NEW TESTAMENT GREEK. _[§ 6¢ 


hardly distinguished even in his day (ὃ 34 (4)). So also ἑστώς neut. 
Mt 2415 χω ; Rev 14! in 046 and a dozen cursives, 


(3) Stems in -n-. 

(a) Medav- black is thus declined : 

Sing. N."wéedas μέλαινα μέλαν; Gi. μέλανος μελαίνης: 
Ῥίι». 1). péraci(v) μελαίναις. 


(Ὁ) An old noun (uéyn) μέγα -- greatness appears to have 
produced the adjective flexion nom. mase. μέγας neut. μέγα 
great, acc. masc. μέγαν (Brugmann (rd? τι. 11. 656). The rest 
of the flexion of this adjective is still in NT supplied by the 
stem peyado-, declined like καλός, which in MGr supplies the 
whole of the flexion, regularly formed (Thumb Handb. 69). 


(4) Stems in -ey- : -u-. 

(a) One very common adjective may be named first, in 
which Brugmann (/.c.) holds that an old neuter noun πολύ 
plenty has produced a nom. masc. πολύς neut. πολύ much 
(pl. many), acc. masc. πολύν. The rest of the flexion is from 
the ordinary stem πολλο- -n- like καλός : the adj. is thus 
parallel altogether to μέγας above, g.v., except that here the 
old irregular flexion survives in MGr (Thumb Handb. 71). 

(Ὁ) Adjectives in -vs form a class which has maintained 
and even extended its ground in ΜΟΥ: see Thumb Handb. 
70 f. Thus decline ὀξύς sharp. . 


Sing. NV. ὀξύς ὀξεῖα ὀξύ ΡΙ. ὀξεῖς ὀξεῖαι ὀξέα 
Α. ὀξύν ὀξεῖαν ὀξύ ὀξεῖς ὀξείας ὀξέα 
G. ὀξέως ὀξείας ὀξέως ὀξέων ὀξειῶν ὀξέων 
0. det ὀξείᾳ ὀξεῖ ὀξέσι(ν) ὀξείαις ὀξέσι(ν) 


In this class the NT shows a few forms each of βαθύς 
deep, βαρύς heavy, βραδύς slow, βραχύς short, γλυκύς sweet, 
εὐθύς direct, θῆλυς female, πλατύς broad (only fem., esp. aS a 
noun = broad way, street), Tpavs unassuming (the Attic mase. 
πρᾷος is obsolete), ταχύς swift, τρᾶχύς rough. 

The Attic gen. sing. in -éos occurs in inferior MSS: βαθέος Lk 24! 
later uncials (τέως SABCDL etc.), πραέος 1 Pet 3: ACP (-éos NBKL). 
For variations in LXX see Thackeray 179. 

The neut. pl. βαθέα in Rev 25: AC 046 is replaced by βάθη NP, which 
is however not a contraction but the pl. of βάθος. 

Nom. sing. neut. θῆλυν occurs in 1) at Mt 191 Mk 10%, and πολύν Ac 


§ 65] ADJECTIVES. 161 


1827, For ἥμισυς (derived from the neuter noun ἥμισυ, which may even 
have survived in popular speech, being very common in the papyri) see 
§ 73. 


IV. Tutrp DEcLENSION. (Two Terminations.) 
(1) Stems inn. 
(a) Stems in -ov- and -ev-. 
These are declined like ἡγεμών and ποιμὴν above (ὃ 56. 1), 


except for the addition of a neuter. Thus σώφρων sober, 
ἄρσην male (appnv—see ὃ 42). 


Sing. N. σώφρων σῶφρον ἄρσην ἄρσεν 
A. σώφρονα ἄρσενα 

Plur. N. σώφρονες σώφρονα ἄρσενες ἄρσενα 
A. σώφρονας ἄρσενας 


In Rev 125 by the usual neglect of concord we have the neut. ἄρσεν 
in apposition to υἱόν : contrast τὸν ἄρσενα v.)°. 


(Ὁ) Comparatives in τῶν (like participles in -#s above) 
combine two stems in their flexion, viz. (-7)-ves- : (-7)-ios- and 
the same in weak gradation with a nasal stem added, -is-on-. 
See Brugmann G'rd.2 τι. 1. 547 ff. for the details! Forms in 
Hellenistic are restricted to those in -οσ- and those in -gov-. 
Thus : 


Sing. Plur. 
N. μείζων μεῖζον μείζονες μείζονα 
and μείζους ( -- -ο(σ)ες) and μείζω ( = ο(σ)α) 
Α. μείζονα μεῖζον μείζονας μείζονα 
and μείζω (-- -ο(σ)) and μείζους (=nom.) and μείζω 
(ὐ. μείζονος μειζόνων 
1). μείζονι μείζοσι(ν) 


(1) The shorter forms occur in Mt 2658, Lk 21° DW, Jn 1°! (exe. δ), 
910, 441, 536 (exc. D), Ac 133! (not D), 1955. 2119, 2313, 21, 2411, 25° (not 
B), 14 

(2) The form in τω has in Hellenistic an indeclinable use, of which 
there are one or two traces in NT. See for this Proleg.50 and Thackeray 
186. So Mt 26° SBD παραστήσει μοι πλείω δώδεκα λεγιῶνας ἀγγέλων, 
Mt 2319 D τί γὰρ μείζω; P Leid C verso #17 (Ὁ. 118—s.c. 160), πλήω 
μου ἔχει χαλκοῦς, P Oxy vii. 1029°4 πλείω τούτων (men) μὴ εἶναι ete. In 
Jn 15° MXA, 53° ABw ete. μείζων, and 210 G ἐλάσσων, we find irrational 
y added to the ace. sing or neut. pl. 


1 As the old mistake by which -ovs was supposed to be contracted from -oves 
is still found in WS 88, the warning against this impossibility is not needless. 


162 A GRAMMAR OF NEW TESTAMENT GREEK. [8 65 


(2) Stems in es. 

These answer to the noun flexion συγγενής in ὃ 58: we 
only have to add neut. sing. συγγενές and pl. συγγενῆ ( = -ε(σ)α). 
There are some sixty adjectives in NT belonging to this class, 
which has however rather strangely vanished in the modern 
vernacular, though that in -vs, which it outnumbers five 
times or more in NT, has survived. See Thumb Handb. 72 f. 


Πλήρης has considerable traces of an indeclinable use, which is often 
found in LXX uncials, and must probably be read in Mk 438 (C¥ cu2— 
it alone explains the variants), Jn 115 (all but D), and—if we must 
follow the MSS—Ac 6° (all but B). Add Mk 8:9 AFGM al., Ac 68 
AEHP al., 1958 AEL 33 cul, 2 Jn§ L. The vernacular evidence will be 
given in full in Vocab. 8.0. : meanwhile see Thackeray 176 f. and refer- 
ences in Prol. 50 n. Despite the rather abundant instances in MSS, 
Thackeray is not inclined to accept the indecl. for the LXX, unless in 
Sir 197° and Job 212: (where a o follows, and the mere transcriptional 
account mentioned above, ὃ 31, may be applied), Deissmann LAE 125-7 
thinks the evidence from papyri early enough to justify acceptance in 
Jn: his view that “in the Gospels and in St. Paul popular forms have 
always a fair claim to preference” leaves us free to exclude it from Ac. 
Then why follow the MSS in one place and not in the other? Can 
Luke have been faithfully copying the popular Greek of his source? 
But it must be admitted that early evidence is local and scanty. P Leid 
C verso #14 (p, 118—s.c. 160) is much the oldest. Next comes a mummy 
label in Preisigke Sammelb. 2632, which Deissmann dates from Augustus. 
P Lond recto 131383 (=1. p. 174—a.D. 78-9) has πληρἾ ace. , pl. This is 
all I know from i/a.pD. and earlier. 


V. MIScELLANEOUS. (Mainly One Termination.) 


The remaining adjectives found in NT only occur in one 
or two forms. 


᾿Αμήτωρ and ἀπάτωρ (only N. sing.) make gen. -opos : 80 ἀπάτορος gen. 
“with father unknown ”—see Vocab. s.v. The neut. would be -op (with 
τέκνον or the like), but I have not seen it. 

“Apraé, nom. pl. ἅρπαγες, ravening. 

Αὐτόχειρ; nom. pl. αὐτόχειρες, with one’s own hand. 

Νῆστις fasting, acc. pl. νήστεις : see above, p. 132. 

Πένης poor, dat. pl. πένησι(ν), might as well be called a noun (stem in 
-r-). Similarly declined is 

Πλάνης wandering, planet (with ἀστέρες), nom. pl. πλάνητες Jud 15 B: 
other MSS have πλανῆται (1st decl.), an alternative form found like 
πλάνης in earlier Greek in this connexion. 

Τετράπους quadruped is only used as a neuter noun, in plural, 


τετράποδα -ων. 


§ 66] ADVERBS. 163 


ADVERBS. 


§ 66. As might be expected, the normal termination τως extends its 
borders considerably in Hellenistic. (It has receded since almost 
entirely : see Thumb Handb. 77.) In NT there are about a hundred of 
these, fully a third of the total number of adverbs (not counting 
adverbial phrases) occurring there. The empiric rule that the adverb 
may be formed by changing the final -ων (-ῶν) of the gen. pl. masc. of an 
adjective into -ως (-ῶς) holds as in earlier Greek. They are attached to 
adjectives of all kinds, including participles used adjectivally (as ὄντως, 
ὑπερβαλλόντως, ὁμολογουμένως, φειδομένως). Νουνεχῶς (from -ἧς adj.) 
may be noticed as replacing the classical νουνεχόντως (in Plato even 
exovtws νοῦν 1) from Aristotle down. Πρώτως (Aristotle) occurs in 
Ac 117° (NBD? and some cursives) for the still far commoner classical 
πρῶτον : cf. P Tebt ii. 2957 and 472 (ii/a.p.), and Rutherford NP 366. 
There is also some tendency to bring -ws into epve. and superl.: thus in 
NT ἐσχάτως (not obviously superl. in form: it is moreover as old as 
Hippocrates), σπουδαιοτέρως, περισσοτέρως. Radermacher (p. 54) asserts 
that the extension of -ws belongs essentially to the written language : we 
might note the appearance of the new adverb ὀλίγως (2 Pet 218) as 
characteristic of the writer’s bookish style—Aquila and the Anthology 
appear to be its only supporters. 

In the vernacular, where as Volker notes! -ws differed less and less 
from -os, the tendency was (as in MGr) to extend the adverbial use of 
the accus. neut. Thus πολλά, πάντα, οὐδέν (whence MGr δέν), ἐξάπινα, 
μέσον, ἐνώπιον. In cpve. and superl. this was traditional : éxtevéotepoy, 
κάλλιον, etc. (see below, § 67), follow classical rule, so do μάλιστα, 
ἥδιστα, τάχιστα, πλεῖστα (papp. passim). 

Other case-endings appear in adverbial use: only a few instances 
need be named, since case-consciousness disappeared centuries before our 
period in nearly all of our exx. The fem. accus. appears in μακράν (sc. 
ὁδόν); ἀκμήν, δωρεάν. Old instrumentals have been recognised in λάθρα, 
πάντη (formed on the analogy of ταύτη, Brugmann Grd.? τι. 11, 713), εἰκῆ, 
πεΐζῆ, κρυφῆ, πόρρω etc. Since there are adverbial datives, like ἰδίᾳ, 
δημοσίᾳ, and the ancient inscriptional witness shows some confusion, we 
cannot be perfectly certain whether to write εἰκῆ or εἰκῇ. See Brugmann’s 
discussion Grd.? τι. ii. 705: it is of course a mere matter of orthography 
for Hellenistic. Forms in -w have no claim to the, and in the rest we 
may leave the matter open. Locatives are recognisable in det (from the 
stem aiuo-, cf. Lat. aevom), πανοικεί, and with short -t (cf. Brugmann /.c. 
710) in the special class in -cori, answering the question in what language ?, 
viz. Ἑλληνιστί, “Popaiort, ᾽᾿Εβραιστί, Λυκαονιστί, Συριστί (in Aramaic) 
ete. Add the old word πέρυσι (see ὃ 106). 


1 Papyrorum gr. Syntaxis Specimen, Ὁ. 9—quoted with approval by 
Radermacher, /.c. 
VOL ἘΠ PAR TR IL rz 


164 A GRAMMAR OF NEW TESTAMENT GREEK. [8 66-67 


Accretions of -s are found in éyyv-s, εὐθύ-ς (orig. neut. acc. sing.), 
μέχρι-ς, ἄχρι-ς (see the chapter on Prepositions), πολλάκι-ς and the 
numeral adverbs. An originally quasi-ablative -@ev has been attached 
in ἐκεῖθεν, πόρρωθεν, πάντοθεν etc., and in a noun (παιδιόθεν), where the 
strengthening with ἐκ (Mk 9?!) reminds us that the original force of the 
suffix is wearing thin. The suffix -dov should be noted in ῥοιζηδόν, 
ὁμοθυμαδόν, σχεδόν. 

For Compound adverbs and Prepositions, see under WORD-FORMATION ; 
and for adverbial phrases the sections on the several cases in the SyNTAx 
(Vol. 111.). 


COMPARISON OF ADJECTIVES AND ADVERBS. 


§ 67. The syntactical developments which have affected 
so seriously the comparison of adjectives and adverbs have 
been described in Proleg. 77 ff. (= Hinl. 120 ff.). They affect 
us at this point by driving out of use many forms that were 
largely employed in the classical period: the mixture of 
compar. and superl. has also produced some strengthened 
forms. There are two forms of comparison, surviving from 
the classical period, one of them of Indo-European antiquity, 
and the other partially so. Adjectives will be given in the 
nom. sing. masc., so that adverbs can be distinguished by their 
ending. 

(1) With suffix -ἴων or yor, -tortos. 
One new formation appears frequently in papyri and Hellenistic 
writers,! and four or five times in NT, viz. τάχειον (better τάχζον), which 
takes the place of θᾶσσον or θᾶττον (= 6ay-1ov—Brugmann, Grd.? τ. 363). 
This fell by its form out of association with the positive and superla- 
tive, and a new form was made on the model βέλτιον : βέλτιστα, κάλλιον : 
κάλλιστα etc. There is no reason whatever for assuming (with WS) that 
a “more original” form emerged for the first time in later Greek, though 
ταχίων and θάσσων might have coexisted (-iyov- as lento-form, -yov- as 
allegro) : where analogy formation explains so easily, we cannot assume 
antiquity without any evidence. The remaining forms are all classical. 
Their obvious shrinking foreshadows their disappearance in later times : 
only κάλλιον survives in MGr (Thumb Handb. 74). 
*Aooov nearer, from ἄγχι (cf. Ger. enger). 
Βέλτιον better, used as epve. of εὖ (elative in 2 Ti 1!8, Ac 1028 D). 
Ἔνγγιστα nearest (Mk 686 Ὁ), sup. of ἐγγύς, which has cpve. ἐγγύτερος 
in class (2). 

᾿Ἐλάσσων lesser, ἐλάχιστος. See below, (3). On the -rr- form see 
§ 43. Cf. Proleg. 236. 

Ἔλαττον less. 


1 Rutherford VP 150; CR xv. 35. 


§ 67-68] ADVERBS. 165 


Ἥδιστα most gladly, sup. of ἡδέως (elative), Ac 138 D, 2 Co 199. 16, 
ens worse. The -σσ- is in all the good uncials. 
“*Hooov worse, less. 
Κάλλιον very well. From καλῶς. See above. 
| Κρείσσων, κρείττων better, κράτιστος. Sup. only as title. For -rr- 
see ὃ 43. The most frequent cpve. of ἀγαθός in LXX. 
Κρεῖσσον, κρεῖττον better. 
Μᾶλλον more, rather, μάλιστα. Positive μάλα not in NT. Sup. 
generally not elative.? 
greater, greatest, μέγιστος. Sup. only once (2 Pet 114, elative, 
as always in Κοινή). 
Μεῖζον more (only Mt 2031), 
Πλείων, πλέων more, πλεῖστος. Sup. only four times, of which two 
| at least are elative : so sometimes πλείων. The forms without ε 
occur in Lk 818, Jn 2115 SBCD (adv.), Ac 1578 (exe. D cu‘), and 
in one or two primary uncials at Mt 57°, Mk 1243, Lk 742-43 1181 
1278 218, Jn 751, 2 Co 26, MGr πιό or myo (from πλεῖον rather 
| than πλέον) takes the place μᾶλλον had in forming comparatives. 
Πλεῖον, πλέον more. 
Τάχιον faster (see above), τάχιστα. Sup. only in Ac 1715, Positive 
τάχα OF ταχέως. 
Ὕψιστος highest. Exc. in the phrase ἐν τ. ὑψίστοις, only used as title 
of God. 


Χείρων worse, used as cpve. of κακός. 


(2) With suffix -tepos -τατος. 
§ 68. The following forms are found in NT: 


(1) Compar.: ἀκριβέστερον, ἀλυπότερος, ἀναγκαιότερος, ἀνεκτότερος, 
ἀσθενέστερος, ἀτιμότερος, βαρύτερος, βεβαιότερος, δεισιδαιμονέστερος, δια- 
φορώτερος, διπλότερος, ἐγγύτερον, ἐκτενέστερον, ἐλαχιστότερος, ἐλεεινότερος, 
ἐντιμότερος, εὐγενέστερος (εὐθυμότερος T.R.), εὐκοπώτερος, ἰσχυρότερος, 
καινότερος, κομψύτερον, μακαριώτερος, μειζότερος, μικρότερος, νεώτερος, περισ- 
σύτερος -ov τως, πολυτιμότερος, πονηρότερος, πρεσβύτερος, πυκνύτερον, 
σοφώτερος, σπουδαιότερος -ws, τελειότερος, τολμηροτέρως OF -ον, τομώτερος, 
ὑψηλότερος, φρονιμώτερος (χρηστότερος T.R.)—see also (3) below. 

(2) Superl. : ἁγιώτατος, ἀκριβέστατος, τιμιώτατος." 


The formations here are altogether upon the old models. 
As may be expected, there are occasional substitutions of 


-ώτερος for -oTepos, and vice versa: the two were equivalent 
long before the later uncials were written, and the traditional 


1 There are places where μᾶλλον appears to be elative, as Mt 6%, In 1 Co 7?! 
we may translate ‘‘ by all means seize (the opportunity)” : cf. Prol. 247. 
2 Overlooked by Blass, p. 33 (also by Debrunner, p. 36). 


166 A GRAMMAR OF NEW TESTAMENT GREEK. [§ 68-69 


distinction meant no more than it does in MGr (cf. Thumb 
Handb. 73). The three superl. in -τῶτος, of which only 
ἀκριβέστατος is true superl., tell of the vanishing of this form, 
which is however still common in the papyri:! Thumb Le. 
shows how it has a limited elative use to-day. Among the 
details note— 


(a) Aurddrepos Mt 9315 is irregular (cf. class. ἁπλούστερος) : it occurs 
in Appian (ii/a.p.) Praef. Hist. Rom. 10 διπλότερα τούτων. The form 
ἁπλότερος occurs in Anth. Pal. vi. 185, and διπλός 2b. x. 101. Cf. Lat. 
duplus, and see Lobeck Phryn, 234. 

(b) On comparative adverbs in -ws instead of -ον see Καὶ. Β]. 1. 577 η.1. 
Both are classical. 

(ὁ) Double comparison.—per(srepos 3 Jn * and its parallels? are best 
explained (like Eng. more, lesser, worser, Ger. mehrere) as efforts to add 
fresh strength to a form the comparative force of which was somewhat 
blunted through its not having the normal termination. MGr π(λ)ειότερος 
and yeporepos, from πλείων and χείρων, πρωτύτερος from πρῶτος, continue 
the tendency. ᾿Ἐλαχιστότερος in Eph 3° is on the other hand, when 
compared with 1 Co 15%, a kind of tour de force in expression, like Aris- 
totle’s τοῦ ἐσχάτου ἐσχατώτερος (Metaph. ix. 4). To this heading practi- 
cally belongs also the frequent use of pleonastic μᾶλλον with cpve. 


(3) Miscellaneous. 


§ 69. (a) A number of comparative formations in -rTepos 
from adverb bases may be put in a class together: 


᾿Ανώτερον higher, from ἄνω. 

᾿Εξώτερος outer, from ἔξω. 

᾿ΕἘσώτερος inner, from ἔσω. 

Κατώτερος lower, from κάτω, Eph 4°. There is a significant altera- 
tion of this to κατώτατα (μέρη) when quoted by sundry Greek 
fathers. 

Κατώτερω lower. 

Περαιτέρω further, from πέραν. 

Πορρώτερον further, from πόρρω. So BA, -pw ND ete. 


1 Asin Hermas (Blass 33, Bl.-Debr. 36): Blass’s suggestion that the Κοινή 
at Rome differed in this respect from that found elsewhere (as in Egypt) is thus 
seen to be unproved. 

2 WS wrongly cft. θεομακαριστότατος Ignat. Polyc. 77: θεομακάριστος is not 
a superl. (see Lightfoot Zgnat. 11. 292). Does this suggest the origin of the 
MGr superl. formations beginning with θεο- (Thumb Handb. 74)? One com- 
pares ἀστεῖος τῷ Oe), which is assumed to be Hebraic. An early ex. in Mim- 
nermus (fr. 119—vii/cent.) ἀμεινότερος. Exx. from papyri in CR xy. 35. See 
Hatzidakis 177, Καὶ. ΒΙ. i. 573. 

3 We have however ἐλαχιστότατος in Sextus Empiricus (iii/A.D.). 


§ 69-70] NUMERALS. 167 


Πρότερος former, from πρό : see below. 
| Ὕστερος later, last. From the adverb which appears in Skt. as wd and 
in English as owt. 
πῶς afterwards, last. 


(Ὁ) Two superlatives are formed with suffix -atos, viz. 
πρῶτος (from *pwFos, Skt. pirva “former”) and ἔσχατος 
(from *éoyos, akin to é€—Brugmann Girt 241). Cf. 
ὕπατος (ἀνθύπατος proconsul) from ὑπό. On the relations of 
πρότερος and πρῶτος see Proleg. 79. 

(c) Many of the substitutes for comparison must be left 
to the Syntax; but one word deserves quoting from MGr, 
περισσότερος as cpve. of πολύς. That περισσός and its epve. 
are practically suppletiva for πολύς is clear in NT. 


NUMERALS. 


§ 70. The following occur in NT: see § 71 (0) for their 
combinations. 


Value. Sign. Cardinals. Ordinals. Adverbials. 
1 ἃ ets one πρῶτος first ἅπαξ once 
2 β δύο δεύτερος δίς 
5. τρεῖς τρίτος τρίς 
4 ὃ τέσσαρες τέταρτος τετράκις 
5 € πέντε πέμπτος πεντάκις 
65 ἐξ ἕκτος 
7 ¢ ἑπτά ἕβδομος ἑπτάκις 
8 ἢ ὀκτώ ὄγδοος 
9 6 ἐννέα ἔνατος 

10 τ δέκα δέκατος 
11 ta ἕνδεκα ἑνδέκατος 
12 ιβ δώδεκα δωδέκατος 
δεκαδύο 
14 ιὃ δεκατέσσαρες τεσσαρεσκαιδέκατος 
15 te δεκαπέντε πεντεκαιδέκατος 


16 ἐξ δέκα ἕξ 
18 τῆ δέκα ὀκτώ or δέκα καὶ ὀκτώ 
Κ εἰκοσι(ν) 


wy) 
o 
mi: 


τριάκοντα 


aS 
Oo 
= 


τεσσαράκοντα 
τεσσεράκοντα 
γι , 
50 @ πεντήκοντα πεντηκοστός 


60 & ἑξήκοντα 


168 A GRAMMAR OF NEW TESTAMENT GREEK. _ [8 70 


70 ὁ ἑβδομήκοντα ἑβδομηκοντάκις 
π ὀγδοήκοντα 
90 9 ἐνενήκοντα 
100 p ἑκατόν 
200 & διακόσιοι 
300 7 Tpraxdotot 
400 ὃ τετρακύσιοι 
500 φ πεντακύσιοι 
600 xX ἑξακύσιοι 
1,000 “a χίλιοι 
2,000 β  δδισχίλιοι 
3,000 γ τρισχίλιοι 
4,000 “ὃ τετρακισχίλιοι 
5,000 “e πεντακισχίλιοι OF χιλιάδες πέντε 
7,000 “€ χιλιάδες ἑπτά ΟΥ ἑπτακισχίλιοι 
10,000 M μύριοι or δέκα χιλιάδες 
12,000 ΜΊβ δώδεκα χιλιάδες 
20,000 ἡ εἴκοσι χιλιάδες or (δισμύριοι---866. below, § 71 ὁ (4)) 
50,000 M μυριάδες πέντε 

100,000,000 μυριάδες μυριάδων. [For the sign see below.] 

’ Those which do not happen to occur in the NT may be supplied. 
Cardinals : 700 Ψ ἑπτακόσιοι--. 800 ὦ ὀκτακόσιοι-. 900 y ἐνακύσιοι. 
Ordinals : 20th εἰκοστός, 100th ἑκατοστός, 400th τετρακοσιοστός (et sim. 

for the rest), 1000th χιλιοστόύς. 

Adverbial : 6 times ἑξάκις, 8 ὀκτάκις, 9 ἐνάκις, 10 δεκάκις (et sim.), 20 — 
εἰκοσάκις, 100 ἑκατοντάκις, 1000 χιλιάκις, 10,000 μυριάκις. 

Since many of these are on the way to become obsolete (see below), we 
cannot expect always to find actual instances in Hellenistic texts. All 
the supplements just mentioned occur in the LXX, except the ordinals 
1,000th and 10,000th and the adverbials 9 times, 100 times, 1000 times, 
and 10,000 times. In the case of the higher numbers this is probably 
not accidental: it is suggestive that we find μυριάκις μύριοι otherwise 
expressed in Rev δ᾽} 918, 


A. Signs. 


Three supplementary signs, drawn from older forms of 
the alphabet, made the available ciphers 27: these were 
στίγμα for 6 (s, in papyri C), κόππα for 90 (φ or 4), σάν or 
σαμπῖ for 900 (4 or T). The signs thus fell into three sets 
of nine each; a@@ units, -- tens, p—T hundreds. These 
horizontal straight strokes (in MSS sometimes curled thus ~, 
as in D) preserve the ciphers from confusion with the letters 
in their ordinary use; but their employment was not 
essential. From 1000 to 9000 the unit ciphers are used 


§ 70-71] NUMERALS. 169 


over again, differentiated in papyri by a large curved flourish 
at the top (74), in MSS by a sloping line below (together 
with the other line sometimes), as ἔ, 5000 (D). The ciphers 
were usually, but not necessarily, placed in order of magnitude 
with the highest at the left, as agi; -- 1916. From 10,000 
upwards the system started afresh, the signs being written 


over a large M, the initial of μυριάδες : thus M = 50,000. 
In printed books ordinary accents are usually employed 
for the horizontal line. 


(a) “In this way the Greeks could express by symbols any number 
less than a hundred millions ... ; and hence perhaps we may under- 
stand why it was that in nations which used this system of notation, the 
next highest number, ten thousand times ten thousand, was used to 
represent a multitude which no man could number, as in Dn 7”, 
Rev 511.71 

(b) The dropping of the horizontal line, which in ordinary arithmetic 
was not needed, made these series of ciphers exactly like words, the 
more so as their order did not matter, and they could be arranged very 
often so as to be pronounceable. Hence no doubt arose the link between 
numbers and names, which on the one side produced mystic words like 
αβρασαξ, the number of the year (since 1+ 2+4100+1+4200+1+60=365), 
and on the other made a name numerically significant, as Ἰησοῦς -- 888. 
For the Greek custom as applied to Rev 1318, see Deissmann LAH 
275-7 : see also Moffatt ΠΟΤ, or Charles ICC, an loc., for the view which 
finds a Hebrew “ gematria” here. 

(c) Fractions could be expressed in words or in the alphabetic nota- 
tion. Sometimes both are found together, as in P Ryl 11. 202 (a) % 10 
(A.D. 108) (πυροῦ) ὀκτὼ τρίτο(ν) iB’, (γίνεται) (πυροῦ) ny if’ “ eight and a 
third and 5}, artabae of wheat, total 8,5, of wheat.” 4 was abbreviated 
4 : neither this nor y nor 6 is found in NT. See Mayser 52 for 
further information. 


8. Cardinals. 

§ 71. (a) Declension and Orthography.—Cardinals are in- 
declinable, except the first four, and those in the table above 
from διακόσιοι onward: these are ordinary plural adjectives, 
while ysAvds and μυριάς are singular collective nouns (gen. 
-d6os). The first four are thus declined: 


N. εἷς μία ἔν δύο τρεῖς τρία τέσσαρες τέσσαρα 
A. ἕνα μίαν ἔν δύο τρεῖς τρία τέσσαρας τέσσαρα 
G. ἑνός μιᾶς ἑνός δύο τριῶν τεσσάρων 

D. ἐν μίᾷ Evi δυσί(ν) τρισί(ν) τέσσαρσι(ν) 


1««The Employment of the Alphabet in Greek Logistic,” by J. G. Smyly 
Mélanges Nicole 519. 


170 A GRAMMAR OF NEW TESTAMENT GREEK. [8 71 


These ‘flexions are followed also when standing at the 
end of a combination, as δεκατεσσάρων. 


(a) Like εἷς are οὐδείς οὐδεμία οὐδέν and μηδείς : for the alternative 
forms (masc. and neut.) οὐθείς μηθείς see § 46. 

(8) The disappearance of the dual flexion of δύο is in line with the 
general development of Κοινή Greek: see Prol. 77 ff. Δύω (pre-classical) is 
found in Ptolemaic papyri by Mayser (p. 313), probably a sheer mistake. 
NT shows neither the gen. δυῶν (Mayser 314) nor the literary Hellenistic 
δυεῖν (late Attic), which appears in some parts of LXX (Thackeray 187), 
as in Josephus, and not seldom in papyri. Avo indecl., a primitive form, 
was ousted from the dative in Hellenistic by the pluralised form δυσί(ν). 
This appears first in Hippocrates, and is traced by Brugmann (Grd.? τι. 
ii. 9) to Ionic. It is regular in literary Kowy from Aristotle’s time, and 
in papyri from the end of ii/s.c.: see reff. in Mayser 314. Δυσίν is of 
course lost in MGr., which has sometimes δυῶ(νε) (Thumb Handb. 81). 
See Vocabulary s.v. for the papyrus record. 

(y) (1) Τέσσαρες is found as accus. extremely often in papyri, and is 
actually normal in LXX (Thackeray 148). We should have expected it 
in ΝΎ uncials, where however it is not adequately attested : see citations 
in Prol. 36 n. WH (App.? 157) make Rev 4* (2nd) the one occurrence 
out of eight where there is not some good authority for τέσσαρες ace. : ef. 
also ib. 138, and Prol. 243, where its predominance in Egyptian business 
documents is noted. It is rather tempting to connect it specially with 
Egyptian Κοινή, in view of its record (a) in LXX, (Ὁ) in papyri and 
ostraca, (c) in &,! the Alexandrian origin of which is becoming more and 
more probable. ΤῸ associate it with the acc. in -es as an element drawn 
from Achaean-Dorian Kowy (as Prol. 36) is on the whole a less probable 
account of its origin than we get by calling in the influence of dvo and 
rpeis, neither of which has a separate acc. form. But both causes may 
have operated. 

(2) Τέσσερα appears in Jn 1925 NALM (not BW), Rev 4° A (not 
NP 046), 48 xA (not P 046), 5° NA, 51#A, 191 RAC: the word itself 
does not occur elsewhere. WH are clearly justified in editing it for Rev., 
on the principle that the MSS are to be followed : it is not so easy to 
admit itin Jn. Τέσσερες and τέσσερας (Rev 4* A) have in any case no 
place in NT, nor in LXX, where τέσσερα is normal (Thackeray 187). 
We cannot therefore regard τέσσερα as Ionic, as the other Lonic forms 
would have accompanied it, recommended further by uniformity. But 
even τέσσερα does not appear in papyri till ii/a.p., and then but rarely.” 
The Ionic sphere of influence, Asia Minor, behaved in this matter quite 
Ἐν ee ee ee ee ee  ΞΞΞΟ ΘΞΕΕΕ 

1 Curiously enough, it is B which shows it in LXX (Octateuch), & only 
having it twice (Thackeray, 73). 

2In OR xv. 331 quoted τέσσερας from CPR 242 (a.p 40), but it must be 
corrected to τέσσαρες acc. Mr. Thackeray also fell into a slip on this matter, 
in his statement (@r. 74) that the form starts in i/a.p. : he admits it ina 
letter to me (Dec. 1910). An Egyptian inser. of i/s.c., in Archiv i. 209, 
no. 22, has δεκατέσΊσερα, and the same word occurs in BGU i. 1839 (a.D. 144-5). 


§ 71] NUMERALS. : 171} 


differently : see Schweizer Perg. 163. If we could assume that a strictly 
localised phonetic change produced epa out of apa, in the period just 
preceding that of δὲ, we might explain the absence of τέσσερας by the 
prevalence of the nominative form (above (1)). The normal acc. thus 
was τέσσαρες, τέσσερα, Which constrained scribes within its area to 
forsake the apa of the LXX and NT autographs. 

(8) Τεσσάρων is invariable in LXX and NT, and τέσσαρσι(ν) is normal 
in both (5 times in NT). But the Homeric and poetical τέτρασι(ν) 
appears in Ac 1011 E Orig, 115 D Epiph, Rev. 208 δὲ (Jud 95: B) : Cronert 
shows (p. 199) it was common in literary Hellenistic. In Rev 7? δὲ has 
the indeclinable form τοῖς τέσσαρες. ΜΟΥ preserves the distinction in 
vocalism between nom.-acc. and gen., the former usually being τέσσερις, 
the gen. always τεσσάρω(ν) (Thumb Handb. 81). 

(8) (1) Between 13 and 19 the forms in the table are firmly estab- 
lished from iii/B.c. down. These were in use as far back as v/B.c. in 
Attic, in places where the substantive preceded the numeral (Thumb 
Hell. 82). In LXX it alternates with the other form (τρεισκαιδέκα etc.), 
but it is almost universal in NT, and MGr has no alternative. Lk 1510 
has δέκα καὶ ὀκτὼ ἔτη (Ὁ) ἔτη τὴ), while ἔτη δέκα ὀκτώ in v.1! happens to 
preserve the classical rule mentioned above: the agreement is mere 
coincidence, as the frequency of such numerals preceding the noun in 
NT shows. δέκα καὶ 6xr®—which is also found in Lk 184 AW, 1: AL— 
is a compromise form found sometimes in LXX (Thackeray 188): the old 
ὀκτωκαίδεκα is reversed by the influence of the cipher τῇ, the commonest 
way of writing.t Whether we should write δεκαοκτώ as well as δεκα- 
τέσσαρες and δεκαπέντε as one word is hard to determine. Ti.-Gregory 
(p. 109) give δεκαοκτώ on the witness of the cursives: WH make an 
exception and print δέκα ὀκτώ, which is perhaps supported by the fact 
that the numeral may be δέκα ὀχτώ as well as dexoyrw in MGr, like 
δεκαννιά or δέκα eva for 19 (Thumb Handb., 80). 

(2) Analogy attempted to extend the rule to 11 and 12. For the 
former δεκαμίαν in an ostracon of ii/B.c. and δεκαμιᾶς in P Oxy ii. 248 
(i/A.D.) are isolated instances. But δεκαδύο flourished during the Ptole- 
maic age, from which Mayser (p. 316) can only cite one instance of 
δώδεκα. Aexadvo appeared in Attica about B.c. 100, and in Asia Minor 
a little earlier. Polybius has it, and sometimes MSS of Josephus.? But 
it died out rapidly, for δώδεκα predominates already in Wilcken’s ostraca 
(Proleg. 246), and MGr has only ἕντεκα and δώδεκα. Wellhausen® states 
about D that “twelve, in the two places where the number is written out, 
is δεκαδύο, not δώδεκα." There are in fact evght instances of δώδεκα in D,* 
with δεκαδύο in Mt 1928, Lk 917, and 18 in 38 places (Lk 84% βι). Aexadvo 


1 The great frequency of this writing in such a MS as D suggests the prob- 
ability that the autographs used symbols rather than words for numbers. So 
in our oldest papyrus scrap (01 in Souter) we find 1d ¢er in Mt 1”. 

2 Meisterhans* 159 ; Schweizer Perg. 164 ; Schmidt Jos. 508. 

3 Hinleitung! 11: ef. Proleg. 96 and Thackeray Gr.188. It does not always 
do to trust in German accuracy without verifying references ! 

4 See the facsimile in Mt 92° 111, Lk 913, Jn 613-6771 119, Ac 197, 


172 A GRAMMAR OF NEW TESTAMENT GREEK. [81] 


appears also in Ac 197 2411 HLPw: Tion the former passage cites very 
scanty cursive witness in six places. As we might expect, δεκαδύο makes 
a rather better show, though still a poor one, in LXX (Thackeray 188). 

(ε) Τεσσεράκοντα is much better supported in papyri than the corre- 
sponding forms of fowr, and in NT the uncials give decisive testimony. 
No single quotation of -ap- can be made from NABC, except Rev 74 A. 
D wavers, having -ep- once and -ap- twice (often j, after its manner). 
W has -ap- twice in Mt 43, and jf elsewhere (ter). The papyri however 
tell a very different tale ;1 and MGr σαράντα, the only form given in 
Thumb Handb. 80, shows that the ε infection was not lasting. It is not 
probable that it appeared in the NT autographs. If they had 4g, as 
suggested above, we can assign the uncial tradition to an age when the 
-ep- form was temporarily established. The fact that forty had the 
syllable before the accent, and four after it, was noted in ὃ 33 as the 
probable cause of the difference between them. 

Note σεράκοντα in Ac 739 C—a first step in the vernacular towards the 
apocopated form of MGr. An ostracon of vi/vii A.D., now in the Rylands 
Library, Manchester, shows the numerals from 44 to 49 written out as 
“ut σερακονταεπτα" etc. A yet more remarkable anticipation was 
developing in the first centuries A.D. Dieterich Unters. 186 cites τριάντα 
from an early inscr., and by ix/A.D. σαράντα and the rest of the series 
were established as they are now. 

(¢) Ἐνενήκοντα in several cursives at Mt 1813 and Lk 15*7 is written 
evvev. : it was natural that ἐννέα should sometimes assimilate the other 
9 forms in which the single ν was correct. 


(ὁ) Arrangement. 


(1) The order which Hellenistic has adopted for the 
"teens 1s kept up usually in τς combinations, as 99 ἐνενή- 
κοντα ἐννέα, 153 (gen.) ἑκατὸν πεντήκοντα τριῶν, 616 
ἑξακόσιοι δέκα ἕξ. In four places «ai is added, as in Lk 186 
above, viz. Jn 22° 5° (where B and minor uncials omit), Ac 
13”, Gal 31%. It is probably not accidental that all five are 
in time-reckonings with ἔτη: there are however exceptions 
enough to make any rule doubtful—thus Lk 2517 13", Gal 21, 
2 Co 122. In Rev 11? and 13° we have μῆνας τεσσεράκοντα 
[καὶ] dvo: the καί is dropped by xP, with C in 13°.  Ptole- 
maic papyri agree with NT usage, with καί proportionately 
rare. The rule was for the numeral to follow the noun 
(Mayser 316 f.). This however does not hold in NT Greek, 

1 Tegoep. seems to appear first in BGU iv. 1105", 1170°, both Alexandrian, 
from about B.c. 10. P Lond 262! (a.p. 11) (=ii. p. 177) has τεσσερακοστοῦ, 
but also recoap. bis. Lists kindly lent me by Mr. Thackeray (completed in 
1906) show -ap- in i/ A.D. 21:2, in ii/ 18:13, in iii/8:7. See § 33 for fuller 
notes. 


§ 71-72] NUMERALS. 173 


where for the earlier numbers the other order predominates : 
from 14 on the numeral more often follows. 

(2) The old method of representing by subtraction num- 
bers ending in 8 and 9 has not survived, except in 2 Co 115 
τεσσεράκοντα παρὰ μίαν forty less one. It is not found in MGr. 

(3) Arithmetical processes are not represented in NT. We may just 
note that an addition sum ends with γίνεται (usually abbreviated) = 
comes to: ef. Ac 44 ἐγενήθη ἀριθμὸς τῶν ἀνδρῶν ὡς χιλιάδες € (Ὁ). Ἐπὶ 
τὸ αὐτό also has an arithmetical connotation,! which suggests itself in Ac 
247, προσθεῖναι ἐπὶ c. acc. may be noted in Mt 6?7=Lk 125 (cf. Rev 
2238), Ἐπί 15 often multiplicative, as χιλιάδες ἐπὶ δεκάδας γίνονται μυριάδες 


(α ἐπὶ 7/M) 1000 χ]10-- 10,000. This use of ἐπί is not unlike Phil 257 
iva μὴ λύπην ἐπὶ λύπην σχῶ. 

(4) It is noteworthy that ycAvoe with numeral adverbs is 
supplanted after 5000 by yeduddes with cardinal: in Ac 4* 
we have this for 5000 as well. It is curious therefore that 
in Ro 114 the ἑπτὰ χιλιάδας of LXX should be deliberately 
replaced by ἑπτακισχιλίους. In MGr χιλιάδες has driven 
χίλιοι out, except for χίλιοι 1000. Just so in NT we note 
that χιλιάς does not come in for the single thousand: it 
begins where it has to be plural. In Rev 145 it is still an 
inflected fem. noun, and in 5!! it keeps its substantive constr. ; 
but in 74% and elsewhere it is already, as in MGr, an adjec- 
tive agreeing with its noun, expressed or understood.’ Simi- 
larly μύριοι gives place to μυριάδες after the unit, but the 
latter retains its substantive character. (So even in late 
papyri, as P Oxy vi. 8961" (a.p. 316) [ἀργυρίου δηναρίων 
μυ]ριάδαν piav.2) In Rev 91° we have it qualified by an 
adverb, as if δισμύριοι; but as it has a dependent genitive 
following it is better to write dts μυριάδες, a noun, or to 
follow x and two or three cursives with δύο. The autograph 


may well have had 8. ΜΜύριοι and -as are obsolete in MGr. 


C. Ordinals. 


§ 72. It is significant that no ordinals beyond /ifteenth 
occur in the NT. ΜΠεντηκοστή was specialised as a feast- 


1[For another explanation of this phrase see Harvard Theol. St. i. 10 ff. 
(C. C. Torrey) ; also discussions in JBL xxxvii. 105 ff. For further treatment 
see the Appendix on Semitisms.—ED. | 

2 Rev 2118 is ambiguous, but 147° makes the adj. more probable. 

3 Cf. for this common combination P Oxy vii. index, p. 256. 


1714 A GRAMMAR OF NEW TESTAMENT GREEK. [δ 72 


name: in papyri it and ἑκατοστή similarly survive, as names 
of taxes. MGr on the same lines uses πέφτη fifth as the 
name of Thursday; but neither this nor the higher ordinals 
are now found, the cardinals supplanting them after τέταρτος, 
This use apparently goes back to Byzantine times, but it does 
not occur in NT’, except in the case of εἷς, That this use of 
εἷς is not due to Hebraism, apart from a modicum of influence 
due to “translation Greek,’ may be inferred from considera- 
tions set forth in Prol. 96. We may add to them the fact 
that the locution εἷς καὶ εἰκοστός or εἰκάς does not go beyond 
the first unit: ἡ μία καὶ εἰκάς is the 21st (day of the month), 
but ἡ ἕκτη καὶ εἰκάς the 26th, 

The ordinals in the ‘teens are formed from the old 
cardinals in which δέκα stands last, after καί. Thackeray 
gives them as “ possibly of Ionic origin.” They were at any 
rate found in Boeotian: thus ἐνακηδεκάτη (Larfeld 16172 
ili/il B.C.), ἑσκηδεκάτη (Thumb Dial. 231) ete. We may note 
that compounds also prefer this older form of the cardinal as 
base: so the series with ἔτος, as ἑπτακαιδεκαετής. What 
NT writers would have used for ordinals between 21st and 
99th we might infer from such forms as δυοτριακοστόν 
32nd (P Ryl 11. 157%&—a.p. 135), τετρακαιεξηκοστόν 64th, 
τεσσερακοσθόγδον 48th. We have also ἑκατοπεντηκοστόν 
150th. 


The type εἰκοστὸς πρῶτος still retains signs of life in LXX (Thackeray 
189), but neither this nor cis δεύτερος καὶ eikoorés—found in Ptolemaic 
papyri (Mayser 318)—appears in NT. In the later LXX books there 
was a tendency to reverse the order, conforming to the normal order of 
the alphabetic signs, which had already affected the cardinals. Thus in 
a petition of iv/a.p. (P Oxy vi. 88917) we find ἑβδο]μηκοστὸν καὶ τρίτον. 
Sundry eccentric forms survived: sy is τετρακαιεικοστόν in early Ptole- 
maic papyri, and is seen still in ii/a.p. (as P Fay 82 and 83) and even 
in iv/A.D. (P Lips 87° τετρακαικοστόν). Ordinals in fractions and days 
of the month were so often written with symbols—as m=the 18th, 
x's'= οὖς etc.—that the papyri present us with relatively few written out 
in full. 

The absence in NT of τετράς, eixds, τριακάς (words of the same class as 
χιλιάς) to denote the 4th, 20th, and 30th of the month is only due to lack 
of opportunity. They were used in LXX as in classical Greek, and in 
papyri well after the NT period. Terpds was transferred to name a day 
of the week (Ps 94 (98) title), and τετράδη to-day means Wednesday. The 
only sign of obsolescence is that they were unused by Theodotion and his 
school (‘Thackeray 189); but the frequency even of τριακάς in papyri— 


§ 72] NUMERALS. 175 


note for example P Oxy vi. 967, a private letter of ii/a.D.—disposes of 
any inference. 

The forms of ordinals that do occur in NT are in other 
respects normal. Ὄνγδοος is nowhere contracted, as (rarely) 
in papyri. Τεσσαρεσκαιδέκατος in Ac 2757. is written 
τεσσαρασκ. in 81, perhaps under the influence of τεσσαρακ,, 
which occasionally appears in papyri. B* writes τεσσαριίσκ., 
as does H the second time: this is found frequently in LXX 
in correctors of B, and once in B*A (Thackeray). It is 
assimilated to τρισκαιδέκατος, an orthographic variant of 
τρεισκ.; but we may remember that téscapis is the normal 
MGr for the cardinal four. In πεντεκαιδεκάτῳ Lk 3! 1, 
drops the -και-. 

The ghost-word δευτεροπρώτῳ Lk 6! (all but p* SBLW, 
some important cursives, and the best versions) will be dealt 
with under Word-composition (§ 104). 


D. Adverbials. 


That in MGr these multiplicatives have disappeared from 
ordinary use—there are survivals like τρίσβαθος “ thrice 
deep,” i.e. “very deep”—makes their fewness in NT easily 
intelligible. There is no sign among NT MSS (so far as Ti 
records) of the forms in -«1, found rarely in LXX and papyri 
(Thackeray 136, Mayser 244): Cronert p. 143 ἢ. gives a 
considerable list of instances from MSS, especially in Josephus. 

In Mt 18” ἑβδομηκοντάκις ἑπτά is in any case abbrevi- 
ated for ἑπτάκις (which is read by D). But the question 
urises whether the -xus has not been added to the wrong 
element: see Prol. 98. W.C. Allen (Comm. zn Joc.) accepts 
the allusion to Gn 4* (first noticed by Tertullian), but 
suggests that in the LXX there and in Mt l/c. we should 
alike translate seventy times seven. In that case the LXX 
mistranslated the Hebrew. Origen took it as 77 times, as 
MecNeile notes in loc. Mr. H. Scott notes the reference in 
the Testaments, Benj. 74, where the phrase is quoted with 
70 x 7 as the meaning. 


E. Distributives. 


Two each, etc., are expressed by δύο δύο, or by ava δύο or 
κατὰ dvo——sometimes the two forms are combined. On these 


176 A GRAMMAR OF NEW TESTAMENT GREEK. [8 72-73 


see Prol. 97; also Thumb Handb. 85 for the corresponding 
usage in MGr. Further discussion is reserved for the Syntax : 
see also the Semitism examined in the Appendix. 


F. Other Numeral Series. 


§ 73. Definite compounds with extant words (such as 
τετράμηνος KTA., διετής KTA.) Will be reserved for the section on 
Word-composition (§ 107); but we may bring in here those 
series in which the numerals are specialised for certain uses 
by agglutinative suffixes—for their history see the section on 
Word-formation by suffixes. 

Abstract numerals, like τριάς triad, do not oceur in the 
NT, except for those like χιλιάς, pupids, which have been 
appropriated for a different purpose: on these see above, B. 
They survive in MGr only in specialised senses, as ἡ Τριάδα 
the Trinity (Thumb Handb. 84). A derivative τετράδιον 
occurs (Ac 12*), meaning a company of four: it is an instance 
of the specialising force of the suffix -voy (Petersen Greek 
Diminutives in -ἰον, p. 84 ff., where τετράδιον is not men- 
tioned). 

Multiplicative numeral adjectives are found with the 
suffixes -πλοῦς (ἁπλοῦς, διπλοῦς, τετραπλοῦς) and -πλασίων (akin 
to our -fold (ἑκατονταπλασίων). It is significant that 30-fold 
and 60-fold are otherwise expressed in Mk 48, though tpsaxov- 
ταπλάσιος (and -wyv) had existed: analogy had clearly ceased 
to be productive. ‘AzAds, διπλός still survive, and διπλός 
makes a new series τρίδιπλος ete. (Thumb Handb. 83). 

Of the n-th day is expressed by a series in -atos: so in 
NT δευτεραῖος, τεταρταῖος, and cf. ὀκταήμερος, which shows that 
the series did not develop in popular Greek. Polybius has 
the regular ὀγδοαῖος, and literary Greek shows a large number 
of these forms. 

Fractions scarcely appear in NT. Τὸ τρίτον 4 (sign γ᾽) 
occurs in Rev 8 sexies, and τὸ τέταρτον 1 (δ in Rev 6% We 
have also the word for half (sign Z), ἥμισυς, in classical Greek 
declined -eva τὸ regularly. It is derived from the proethnic 
semi- (ἡμι-, Lat. s¢mi-, O.K. sam (in Shakespeare’s sandblind) 
with the suffix -tw. ‘“Hye- occurs in many compounds, as 
ἡμίωρον half an hour, ἡμιόλιος one and a half (adj.). Some 
curious features arise in the spelling and declension of ἥμισυς 


§ 73] . NUMERALS. 177 


in NT times. The fem. ἡμίσεια has been cut loose from the 
declension, and is only used as a noun (80. μοῖρα); while 
ἥμισυς, as an adj. of two terminations, or even as an indeclin- 
able quasi-numeral ἥμισυ, is almost entirely confined to the 
forms ἥμισυ (ἥμισου) and (much less frequently) ἡμίσους and 
ἡμίσει. 


(a) The spelling ἥμυσυς -v is common between iv/ and i/B.c., the 
second syllable being assimilated to the third: very rarely it is carried 
into forms where -v does not form part of the suffix ‘In the Ptolemaic 
papyri this form predominates in iii/B.c., in 111 B.C. ἥμυσυς and ἥμισυς 
are represented by nearly equal numbers” (‘Thackeray 95): he adds that 
the absence of ἥμυσυ from the LXX is unfavourable to the trustworthiness 
of the uncials.! Its absence from the NT will, on the same showing, be 
a good sign, for with the rapid movement of v towards the simple 7-sound 
ἥμυσυ became obsolete : only six instances can be cited from the imperial 
age by Cronert, and in NT it is represented only by ra ἥμυσοι (-o1=-v) in 
Lk 198 D*—indecl., with ἡμίση in correction : see ὃ 35. 

(b) Another peculiarity is thus noted by WH (App.? 165): “In Ap? 
ἥμισυ each time has the v.l. ἡμίσου (A*, SA, X*: cf. Is 4416 B), which 
likewise is one of the variants for ἡμίσους Mc.” In Mk 67° LAW read 
ἕως ἥμισυ, and it seems better to regard ἡμίσου (? ἥμισου) there as the 
same reading: for this curious form see (6) below. The indecl. ἥμισυ 
appears also in Lk 198 τὰ ἥμισυ AD* (ημυσοι) RA 69: cf. Tob 1910 B. 

(c) Ta ἡμίσια Lk 198 SB*Q 382 and L (-eca) may be supported by a 
Pisidian inscr. of the imperial age (Papers of Amer. School 111. 204), whence 
Crénert cites ἡμυσίοις. It is obviously useless to cite fem. forms in -σια 
(with WH). W. F. Moulton (WM 75 n.") compares ὀξεῖα χρέμισαν in 
Hesiod Sc. 348, and θήλεια neut. in Aratus 1068: WS 87 adds from 
Meisterhans a neuter πλατεῖα dated B.c. 358. Thackeray 179 gives LXX 
parallels. MGr, which has developed the adj. in -vs, makes plur. βαθειοί 
-ἔς -d, which is the descendant of these forms. 

(d) The older forms of the gen. do not appear in NT uncials, except 
for ἡμίσεως Mk 655 S and ἡμίσεος ab. ΤΙ. 

(e) Some account of general papyrus usage may be added, drawn 
mainly from Mayser (294 f.), supplemented from an analysis of a large 
number of documents dated A.D., containing about 90 occurrences. In 
these last no plural occurs—the Hellenistic ἡμίσεις and ἡμίση are barely 
quotable a.pD.—and ἡμίσια (so spelt, 8 times) is only used asa noun. (It 
is absent altogether in LXX.) Τὸ ἥμισυ, with or without μέρος, increases 
in frequency. Ἡμίσους gen. and (less frequently) ἡμίσει dat. occur freely, 
and may agree with fem. nouns, as ἀρούρη[ς ἡ μίσους BGU ii. 42212 (ii/a.p.). 
Ἥμισυ (-ov) indecl. is in these documents restricted to the position 
following an integer, as ἀρταβῶν ἑβδομήκοντα δύο ἥμυσυ BGU ii. 53833 
(100 a.p.), another hand having already written the same phrase with 


1 To his one ex. (Dn Θ 7” B) Cronert adds Nu 2813 in a palimpsest edited 
by Tischendorf; also one from Cod. G of the Octateuch (iv/v A.D.). 


178 A GRAMMAR OF NEW TESTAMENT GREEK. [§ 73-74 


ἡμίσους. Evidence for its wider use may be seen in Thackeray p. 180. 


A.D.: add a papyrus cited by Crénert (Mem. 23) with οἵμυσου (A.D. 261), 
and NT uncials as above. 

(f) The MGr μισός is prepared for in the Doric ἥμισσος (Syll. 594 bis, 
598°, both ii/B.c.), and ἡμίσωι IM Ae iii. 168 (i/B.c.), Syll. 493", τὸ ἥμισον 
Syll. 5967 (ii/B.c.): we might even accent this derivative oxytone, as in 
MGr. 

(9) No instance can be quoted in which καί links ἥμισυ with an 
integer preceding. It is natural therefore that 046 and a good many 
cursives should omit the irregular conjunction in Rev 119 ἡμέρας τρεῖς καὶ 
ἥμισυ : it is significant that in v.11, where the order is changed to τρεῖς 
ἡμέρας καὶ ἥμισυ, only two cursives omit. This is of course only one more 
irregularity added to the special grammar of this Book: we need not 
suspect the reading. Note that the idiomatic use of ἥμισυ indecl. follow- 
ing an integer survives in MGr, as δυό ᾽μισυ 24, τρεῖς ἥμισυ 34 (Thumb 
Handb. 82). 


PRONOUNS. 


§ 74. The Pronoun system retains one or two special 
flexions which from prehistoric times differentiated it from the 
Noun. Chief among them is the nom. ace, neut. sing in -ο(δ), 
with which we compare the form surviving in Skt. tad, Lat. 
istud, Eng. that, ete. 

Demonstrative Pronouns include ὁ (the Definite Article), 
οὗτος this, ὅδε this, ἐκεῖνος that, τοιοῦτος such, τοσοῦτος so much, 
τηλικοῦτος 80 great, τοιόσδε such, 


Thus decline 


Sing. N. ὃ ἡ τό οὗτος αὕτη τοῦτο ἐκεῖνος ἐκείνη ἐκεῖνο 
Α. (566 8 48) τοῦτον ταύτην τοῦτο and the rest like καλός 


σ. τούτου ταύτης τούτου 
1. τούτω ταύτῃ τούτῳ 
« t t 
Plur. N. οὗτοι αὗται ταῦτα 
vA. τούτους ταύτας ταῦτα 
σ. τούτων 
2. τούτοις ταύταις τούτοις 


"Ode (rare) is declined as ὁ with enclitic δὲ added: 
similarly τοιόσδε is τοιός like ἅγιος with δὲ (only once found 
2 Pet 1). The others are like οὗτος : eject the initial τ and 
prefix tov-, τοσ-, THALK- respectively. 


They may however take -ον in nom acc. sing. neut.: thus τοσοῦτον 
Heb 12}, τηλικοῦτον 2b. δὲ δ, but τοσοῦτο Heb 722 S*ABCD*P 33 cu’; in 


§ 74-76] PRONOUNS. 179 


Mt 18° τοιοῦτο NBLW al., τοιοῦτον D etc. In Heb. Blass might have 
regarded it as significant that τοσοῦτον precedes a vowel, and τοσοῦτο a 
consonant. Both are inherited from earlier Greek. 


“AdNos -n -o (otherwise like καλός) other, and ἕτερος (like 
ἅγιος) different (sometimes still other of two). 

‘O δεῖνα so-and-so (A τὸν δεῖνα, G τοῦ δεῖνος, D τῷ δεῖνι) 
occurs once (in ace. Mt 2018). 

“Exaotos each (like καλός). 


§ 75. Relative Pronouns are ὅς and ὅστις who (see 
SyNTAX), ὅσ-γε and ὅσ-περ (1.6. ὅς with enclitics—see Synrax), 
οἷος of which kind, such as, ὅσος as many as, as much as, ὁποῖος 
such as: the last three are declined like καλός or ἅγιος 
regularly. 


Interrogative Pronouns are τίς who?, ποῖος of what 
sort ?, πηλίκος how great?, πόσος how many? (like ἅγιος 
and καλός). 

The Indefinite Pronoun is tus some, any (enclitic). 

Thus decline: 


Sing. NV. ὃς ἥ 6 ὅστις WS OT) > TIS τό τὶς τὶ 
A. ὃν ἥν ὅ (6s with enclitic ta τί τινά τι 
G. etc., like ὅτου Tis) τίνος τινός 
D. Article, τίνι τινί 

Plur. N. but accented οἵτινες αἵτινες ἅτινα τίνες τίνα τινές τινά 
A, and without τ. τίνας τίνα τινάς τινά 
Ge τίνων τινῶν 
D. τίσι(ν) τισί(ν) 


“Ὅστις is only used in nom. (and acc. neut.), apart from 
the old additional gen. neut. ὅτου, surviving in the stereo- 
typed phrase ἕως ὅτου (as far as what) until, and in Lk 13” 
ἀφ᾽ ὅτου D. Its oblique cases are rare in vernacular Κοινή: 
sometimes we have them with additions equivalent to our 
-soever, aS ὁντιναδηποτοῦν whomsoever. The neuter ὅτι is 
often (very needlessly) printed 6 τὰ or even ὅ,τι to distinguish 
it from ὅτε that. 


§ 76. Personal Pronouns are ἐγώ J, ov thou, αὐτός -ἡ -ὁ 
he, she, τέ. 
VOL. Il. PART Il.—13 


180 A GRAMMAR OF NEW TESTAMENT GREEK. [§ 76-77 


Thus decline: 


Sing. WN. ἐγώ N.V. σύ N. αὐτός -ή -ό 
A. ἐμέ με σέ σε αὐτόν -͵ν -ό 
G. ἐμοῦ μου σοῦ σου αὐτοῦ -ἧἢἣς -οὔ 
1). ἐμοί μοι σοί σοι αὐτῷ -ῇὟ τῷ 

Plur. N. ἡμεῖς N.V. ὑμεῖς N. αὐτοί -al -ά 
A. ἡμᾶς ὑμᾶς αὐτούς -ἄς -ά 
G. ἡμῶν ὑμῶν αὐτῶν 
1). ἡμῖν ὑμῖν αὐτοῖς σαῖς -οἷῖς 


For the use of the enclitic forms see the Synrax. After 
prepositions the enclitic forms are not used, except with πρός 
(generally). 

Αὐτός is properly demonstrative, and has an adjective use 
= self, same, for which see SYNTAX. 


§ 77. The Refiexive Pronouns are ἐμαυτόν -yv myself, 
σεαυτόν -nv thyself, ἑαυτόν -nv -ὁ (rarely αὑτόν -nv -0) himself, 
herself, itself. Thus decline: 


Sing. A. ἐμαυτόν -ἦν σεαυτόν -ἦν ἑαυτόν -ν -ό 
G. ἐμαυτοῦ -ῆς σεαυτοῦ -ἧς ἑαυτοῦ -ῆς -00 
3 Le ~ an σι € ~ a A 
1). ἐμαυτῷ -ἢ σεαυτῷ -ἣ ἑαυτῷ -ἢ -@ 
Plur. A. ἑαυτούς -as -ά 
G. ἑαυτῶν all persons 


1). ἑαυτοῖς -ats -ots 

For the vernacular spelling ἐμᾶτόν see ὃ 37. 

The shortened forms σαυτόν and αὑτόν occupy a somewhat 
ambiguous position. The former is non-existent in NT, 
except for one appearance in B (Jas 25) and one in a few 
cursives (Rom.14”). But the latter is read by WH in some 
twenty places, and the strength of their case seems irresist- 
ible: in Jn 224 αὐτὸς δὲ ᾿Ιησοῦς οὐκ ἐπίστευεν αὑτὸν αὐτοῖς, 
and Lk 23” προὐπῆρχον γὰρ ἐν ἔχθρᾳ ὄντες πρὸς αὑτούς, it 
is simply impossible to read the ordinary demonstrative. 
The ἢ was faint at this time, but still heard; and in spite 
of serious difficulties it seems imperative to believe it could 
sometimes differentiate the pronouns as in much earlier 
Greek. See Vocabulary s.v. αὑτοῦ. 

The difficulty lies mostly in the fact that our outside evidence proves 


overwhelmingly the disappearance of the dissyllabic forms of 2nd and 3rd 
person reflexives before the age of the NT. In Attic inserr. éaur. has to 


§ 77-79] PRONOUNS. 181 


aut. the ratio 31 : 23 from B.c. 403 to 300, while from 300 to 30 it rises to 
100 : 7 (Meisterhans? 153). In Egypt, in the Ptolemaic inserr. and papyri 
included in Mayser’s survey (p. 305f.), avr. outnumbers éaut. by 3:1 
in i1i/B.c., the proportion is reversed in ii/B.c., and in i/B.c. éaur. stands 
alone. In Pergamon during the pre-Roman period ἑαυτ. stands at 24 : 5 
above air., while under Roman rule it rises to 18 : 2 (Schweizer Perg. 
162). In Magnesia no certain instance of air. is forthcoming (Nach- 
manson Magn. 144). These statistics suffice to show that air. was very 
“near extinction before a.p. But as we look at Mayser’s analysis for the 
second half of the Ptolemaic period, we find that the large majority 
against avr. is secured by official papyri and inscrr.: in private docu- 
ments there is absolute equality. A priori we should expect to find air. 
vanish first in a country where psilosis was complete, since it would no 
longer be distinguishable from air. This is true of Asia Minor, as 
Pergamon and Magnesia attest. But the h may have survived elsewhere, 
and the actual citations that are given from Egyptian documents show that 
avr. had not yet ceased to make sporadic appearances, Thackeray (Gr. 
190) shows that αὑτοῦ still exists in LXX, though scantily). Against 
Blass’s denial, we must leave room for the possibility of very occasional 
retention of the dissyllabic form. See Vocab. s.v. ἑαυτοῦ for post- 
Ptolemaic instances of αὑτοῦ. 


For the plural of the reflexive, ἑαυτούς stands alone in 
the Hellenistic age. A survival of ὑμῶν αὐτῶν may be noted 
in 1 Co 5%, which is quoted from the Pentateuch: in this 
part of the LXX (Thackeray 191) these forms were still 
current. On ἑαυτούς as the common plural for all persons, 
coupled with the exclusion of ἑαυτόν from 1st and 2nd 
person, see Prolegy. 87. 1 Co 10” has τὴν ἑαυτοῦ where 
τὴν σεαυτοῦ might have stood (so D,*); but we may render 
impersonally “ one’s own.” 

Only negligible MSS violate this rule, except in Jn 1894, where A 
and W join the crowd with ἀφ᾽ ἑαυτοῦ “of thyself,” and Jn 1433 ἐμφανί- 
ζειν ἑαυτόν, read by three Ferrar cursives and therefore presumably their 
archetype. In Mk 1** W has δεῖξον ἑαυτόν, and in Lk 2359 actually 


σῶσον καὶ αὑτὸν καὶ ἡμᾶς. The LXX citation in Lk 1027 has ὡς ἑαυτόν 
1h 
in A, as elsewhere in inferior uncials. See Thackeray 190 f. 


§ 78. The Reciprocal Pronoun is ἀλλήλους -ων -o1s : no 
fem. or neut. forms occur in NT. ‘Eavtovs is also used, and 
sometimes phrases with ἄλλος or eis—see the SYNTAX. 

Possessive Pronouns, attached to the Personal, are ἐμός 
my, σός thy, ἡμέτερος οἱ", ὑμέτερος your, all declined as 
regular adj. “Idvos own belongs to the same category. 


§ 79. Pronouns indicating duality, as πότερος which of 


182 A GRAMMAR OF NEW TESTAMENT GREEK. [§ 79-80 


two?, ἑκάτερος cach of two, against τίς which and ἕκαστος 
each, are obsolete in the vernacular: see the Synrax. The 
NT has only ἀμῴῳφότεροι both, and ἕτερος other; but the 
former has begun to lose its duality, and the latter has 
almost entirely lost it—-see Prol. 79 f. and further in SynrTax. 
For this place also will be reserved tables of Correlative 
Pronouns and Pronominal Adverbs. 


VERBS. 


§ 80. The verb in Hellenistic Greek has been simplified 
in many directions, as compared with earlier profusion of 
forms; but except for the complete elimination of the Dual 
no category has been definitely removed. MGr has entirely 
lost two Moods, the Optative and (except dialectically) the 
Infinitive. It has also treated the Middle Voice and the 
verbs in -we just as Latin treated them in prehistoric times. 
A few survivals serve as exceptions to prove the rule. These 
and other tendencies, the issue of which is seen in MGr, were 
all at work early in Hellenistic; but they had’ not travelled 
far enough to relieve the accidence of much grammatical 
lumber, once significant but now outworn. A brief summary 
may be given before we present the paradigms. 

Conjugations.—For practical purposes the verb may 
still be divided into the familiar categories of Simple -w 
Verbs, Contract Verbs, and verbs in -μο The last-named 
are being largely replaced by forms of the other two classes ; 
and among the Contract Verbs there is a tendency towards 
the fusion of -aw and -éw forms, which however has not yet 
gone far. 

Voice.—The Middle and the Passive have drawn closer 
together in form, while Active endings have replaced a good 
many Middle where there was no clear distinction of function. 

Mood.—The Optative has very largely disappeared, 
being restricted to a few uses; but for a semi-literary pre- 
dilection in the Lucan writings, we might leave it out of the 
paradigms and merely set down isolated forms. 

Tense.—The Strong Aorist survives (in one or more 
Voices) in less than seventy verbs, and most of its occur- 


§ 80-81] VERBS. 183 


rences are accounted for by the commonness of a verb’s use 
preserving ancient forms. The Weak Aorist is constantly 
encroaching ; and its endings are steadily driving out those 
special to the Strong Aorist, even in the verbs which keep 
the old root form. Weak Aorist endings moreover tend to 
oust the proper suffix of the 3rd plural Perfect Active, and 
so complete the identity of person-endings between these 
tenses. In past tenses of the Indicative a weakening of the 
Augment’s hold has begun in compound verbs. 


CONJUGATION AND TENSE STEMS. 


§ 81. The complexities of the Greek verb are due mainly to 
the survival of conjugation stems, which give great variety to 
the present tense and its attendantimperfect. In prehistoric 
Indo-Germanic these stems may have carried some functional 
distinctions; but it is difficult to prove these distinctions in 
all cases, and most of them were obsolete before Hellenistic 
Greek arose, even if they could be claimed for earlier stages. 
A brief sketch of the historical classification of present stems 
may be given, so far as concerns words occurring in NT: for 
a full account reference may be made to Brugmann-Thumb 
Gr. 316 ff., or to Giles 425 ff. from which the numeration of 
the classes is taken. 

The primary division is that between Thematic and 
Unthematic formations, which accounts for the obvious 
classes of -w verbs and -μὲ verbs, and for other peculiarities 
of LE. verb-systems. Thematic formations show the vowel o 
in the 1 sing. and plur. and 3 plur. (exc. imper.) and ε in 
2 sing. and plur. and 3 sing. In unthematic formations the 
person-ending is added directly to the root or the tense- 
stem. We need not here discuss whether the ὁ : 6 is histori- 
cally a part of the root, ejected in unthematic formations by 
the prehistoric action of accent, or a functional suffix: all 
these questions belong to a period which was over for Greek 
ages before Homer. The study of the Hellenistic verb soon 
shows that unthematic formations were receding fast before 
thematic, which in MGr cover the whole field except for the 
substantive verb. This moreover is of Middle form; and 
Hellenistic largely retains unthematic Middles where the 
Active has become thematic. 


184 A GRAMMAR OF NEW TESTAMENT GREEK. [8 81 


A, Conjugation Classes, Present Stem. 


The following are the conjugation classes: in each case 
(a) is thematic, and (6) unthematic. Where no unthematic 
forms survive in Hellenistic, no distinction is attached. 


I. Person suffixes are added to the root— 
(a) with thematic vowel. 
Thus édvo-v, édve-s, λυό-μεθα, λύε-σθε. 
(b) without thematic vowel. 

Thus ἔσ-τι, pl. εἰσί (for ἑντί, ie. sentt) ; 

ἔφη-ν, pl. ἔφἄ-μεν, mid. ἐφἄ- μην. 

The singular active had strong root-form, the 
plural and the whole middle had it weakened : 
thus én : dd, ec : o. Here levelling has 
obliterated much of ἃ distinction which 
served no special purpose. 


Ta. Reduplicated forms (with ὁ in redupl. syllable). 
(a) thematic. 
Thus πί-πτ-ο-μεν (J pet), t-C-e-re (= si-2d-, 
J sed). 
(6) unthematic. 
Thus ἕστη-μι, pl. ἵ-στἄ-μεν, mid. ἵ-στἄ-μαι 
τί-θη-σι, pl. Te-Oé-acr, mid, ἐ-τι-θέ-ντο. 


II. With formative suffix in -n-. 
(a) 
(a) suffix vo : ve or avo: ave. 
(i.) added to root. 
Thus αὐξ-άνο-μεν, τέμ-νε-ται. 
(ii.) added to root with a nasal inserted (only 
-dvo). 
Thus Aap-Bdvo-pev (AaB-etv), ἐχάνθ-ανε (λαθ- 
εἶν), Nayy-ave-Te (λαχ-εῖν). 
(0) suffix νη : νᾶ added to root. 
Thus δύ-νἄ-ται. 
(8) 
(a) (i.) suffix vuo-: vue. 
Thus δεικ-νύε-τε etc.: verbs in -vume from 
classical times thus tended to become 
thematic. 


ὃ 81-82] VERBS. 185 


(ii.) suffix νέο : vFe. 
Thus φθά-νο-μεν, Ti-ve-Te. 
(iii.) suffix vero : vere. 
Thus κινούμεθα (=-verd-weba), ἀφικνεῖτο 
( = -véFe-T0). 
(0) suffix vv : vd (from new : nw) added to root. 
Thus δείκ-νῦ-μι, ἐδείκ-νύ-τε, δεικ-νὕ- μεθα. 


III, With formative suffix in so: se. 
Thus κλά-(σ)- ὦ, omd-(c)-@ av&-w ( = aug-so-). 
IV. Stems in sko : ske. 


The suffix is added to simple or reduplicated stems, 
sometimes with ὁ before it. 
Thus βό-σκω, εὑρ-ίσκω, δι-δά(κ)-σκω, γι-(γγνώ-σκω. 


V. Stems in to : te. 


This class may be ignored for our purpose. Verbs in 
-7Tw do not belong to it: see VII. 


VI. Stems in Oo : Oe. 


Thus πρή-θω, ἔσ-θω : it is very small. 


VII. Stems in yo : ye. 


This is a very large class, varying greatly in form accord- 
ing to the consonant or vowel preceding. A large proportion 
of these verbs are formed from nouns. Among the principal 
types are: 


(i.) with yo: ye added to root. 
Thus Baw (= LE. g"m-io-), χαίρω (-7-i0-). 
(ii.) added to another suffix. 
Thus κρίνω (kpiv-yo: κριν = κρῖ-ν). 
(iii.) added to the stem of a noun. 
Thus φυλάσσω (xy), ἀστράπτω (Ty), ποιμαίνω 
(ny), ἐλπίξω (dy), μαρτύρομαι (Upy), τιμάω (ay) 
fw, φιλέω (ey), μεθύω, ete. 


TENSE STEMS. 


§ 82. These are essentially of the same nature as the con- 
jugation stems; but having developed definite functions, they 


186 A GRAMMAR OF NEW TESTAMENT GREEK. [Ἃ8 82 


came to be formed from roots belonging to any one, or more 
than one, of the stems given above. 


B. Strong Aorist. 


This is a special use of formations already described 
under A. I. In the indicative it has of course only the 
augmented form. 


(a) =I. (a) with weak gradation in the root. 
Thus ἔλιπο-ν (γ,λειπ), ἐσπάρ-ην (γίσπερ, with rr 
in weak gradation, and passive suffix), ἔπαθο-ν 


(νπενθ, with a=), ἔσχο-ν (4/cey). 


Sometimes the rule of weak gradation is broken, when 
the present stem is differentiated by the formative of another 
class. Thus αἰσθέ-σθαι from αἰσθ-άνο-μαι, ἔτεμο-ν from τέμ-- 
vo, ἔπεσο-ν (for ἔπετον) from πί-πτ-ω, εὗρο-ν from εὑρ-ίσκω, 
ἐγενό-μην from γίνομαι ( = γι-γν-). 


(b=): 
Thus ἔστη-ν, ἔθε-μεν, δό-σθε, γνῶ-θι. 
The Strong Aorist in the passive is not thematic, being 
formed with an 7 which has no connexion with the thematic 
vowel. 


C. Weak Aorist. 


The stem is formed by adding o to the root, with the 
suffixes attached directly. Forms answering to this descrip- 
tion have disappeared from Greek, except for the Ist sg. 
(ἔδειξα = édeiks-m), and in active and middle, apart from the 
subj. and two or three other forms, the characteristic of the 
tense is ga. After liquids this o is obscured. 


Thus ἔλεξα (γ΄λεγ), ἐλυσά-μεθα (4/Av), ἔκρινα (xpi), 


ἔσπειρα (4/omrep), ἔνειμα (ψίνεμ), ἔκτεινα (γίκτεν), 
ἔστειλα (γ΄ στελλ). 


The Weak Aorist in the passive is formed in a wholly 
different way. A new formative θη, drawn originally from 
a middle person-ending (2nd sing.) -@ys, has extended 
through the whole tense system, with endings following those 
of the Strong Aorist. 

Thus ἐδό-θης, κρι- θῆ-ναι. 


§ 89] VERBS. 187 


1). Future Stems. 


The Future is partly extended from the Weak Aorist 
subj., which before the analogy of the pres. subj. of thematic 
verbs affected it had the flexion λύσω -εἰς -εὐ -opev -ετε -ουσι, 
and similarly in the middle. Probably there was also some 
influence from a special future stem in syo: sye, found in 
Aryan and Lithuanian, which coincided in form. In most 
verbs accordingly the Future act. and mid. coincide in stem 
with the Weak Aorist, but keep -ow when the aorist has a 
special form, as δώσω (aor. ἔδωκα). 

In verbs with Liquid stems, and in a few others which 
may be sought in the Table, the Future stem is formed with 
-ἐ(σγω instead of -ow, and a flexion results identical with 
that of the Contracta in -éw. Thus from κρὶν (xpivé(c)w) 
κρινῶ, from σπερ σπερῶ, from ἐλπίζω ἐλπιοῦμαι. 

In LXX and occasionally in papyri and inscriptions we find a similar 
future from stems in -d(w: thus ἐργᾶται from ἐργάζομαι. The absence 
of this in NT is marked, and shows a dialectic distinction: thus in 
1 Co 216 the LXX συμβιβᾷ is altered to συμβιβάσει. The formation is 
found in Ionic: see Thumb Dzal. 358. 

In the Passive the Future is formed from the (Strong 
and Weak) Aorist stems by adding -cowae to the character- 
istic (θ)η. Thus κριθήσομαι, ἀνοιγήσομαι. The “Third” 
Future, or Future Perfect, is obsolete; but the form Kxexpaé- 
ομαι (from Kéxpaya) is on the same model. 


E. Perfect Stem. 


This stem is unthematic, and in the earliest period (as in 
classical Sanskrit, and in our own speech-family up to the 
Middle English period) had vowel gradation, with strong 
root in singular active and weak elsewhere. So in classical 
Greek οἶδα, 2 pl. ἴστε, just as in Chaucer’s English J wot, we 
witen. In Hellenistic this verb is assimilated to other 
perfects; and the only trace of gradation is that the middle 
stem sometimes differs from the active. 

The Perfect stem (except in the verb οἶδα) is redupli- 
cated, with e in the reduplicating syllable. It has in the 
Active two formations, Strong, with internal change as in 
our own Strong Perfect, and Weak, with a formative suffix κ. 
Thus λέλουπα from λείπω is historically parallel with our 


188 A GRAMMAR OF NEW TESTAMENT GREEK. [§ 82-83 


rode from ride. Roots with radical e show o in its place in 
the Strong Perfect active, except when combined with ὅν 
(πέφευγα from φεύγω). The suffix « only appears when the 
verb stem ends in a vowel or a dental mute: thus λέλυ-κα 
from Av-w, πέπει-κα from πείθω. 

The past tense of this stem, called Pluperfect, has 
(usually) the augment, and a special set of person-endings 
with the connecting vowel e, which in Hellenistic goes right 
through. : 

The Middle and Passive have one set of forms, Perfect 
and Pluperfect, which are normal unthematic forms from the 
Perfect stem. 


F, Verbals. 


Two adjectives are formed from verbal roots, unconnected 
with the tense system. One in -τός, historically identical 
with the Latin perf. partic. passive in -tws (-sus) and our 
English -d participle, is extended to ‘derived verbs and 
attached to their stem: the form can usually be deduced 
empirically by putting -ros for the -σω of the Future. For 
the function of this verbal adjective, see Proleg. p. 221 f. 

The gerundive in -τέος, formed in the same way, is very 
rare in Hellenistic (only once in N'TT—Lk 555). 


VERBS. 
AUGMENT AND REDUPLICATION. 


1. Augment with 7. 

§ 83. The augment with 7 is found in later Attic (since 300 B.c.) 
in the verbs μέλλω, δύναμαι, βούλομαι, probably by analogy of ἤ-θελον, 
where 7)- was a preposition (Lat. @, Skt. α).} Of these forms ἠβουλ. 
never occurs in NT (exc. 8 in Philem 18, HLP in Ac 2818 and cursives 
in 2 Jn 13), though sometimes in LXX, and once in a i/A.D. papyrus 
(P Oxy ii. 281!°). So also in patr. often, esp. in aor. Ἤμελλον and 
ἔμελλον alternate: WH print ἤμ. always in Lk and Ac, exc. Ac 2157, 
but in Jn both forms about equally. Avvaya has aor. 76. (often with 
variant ¢6.), but great fluctuation in the impf.: 76. is read by WH 7 
times and ἐδ. 12 times. See on both MG sv. and WH App.? 169. 
Θέλω (never ἐθέλω) makes 76. always, as in class. Gr. To the same class 
are generally assigned ἑώρων, dvéwéa (and other forms from ἀνοίγω), 
€eaynv, in which by “quantity metathesis” yo etc. would become ca, ew 


1 See for its extent Meisterhans* 169 ; for its origin, Brugmann Grd. 11. ii. 
§ 634, Giles? 408. It appears occasionally in earlier poetry : see Veitch s.vv. 


§ 83] VERBS. 189 


ea.1 ‘Edpov only Jn 6? NT etc. where ἐθεώρουν BDL(W) is the better 
reading. ᾿Ανοίγω (διανοίγω), as the simple verb became obsolete, took an 
augment in the preposition, either with or without that in the root 
syllable : hence the triple types nvew. (with inf. ἀνεῳχθῆναι), nvo., avew. 
For the distribution of forms between the types see WH App.” 168. 
Kardyvupe has aor. κατέαξα, pass. -eaynv, but in the latter the augm. is 
continued in subj. κατεαγῶσι Jn 19%! (as in ἀνεῳχθῆναι above). Veitch 
quotes κατεάξαντες from Lysias and κατ-εαγῇ -εαγείη -eayeis from the 
Ionic Hippocrates. A more difficult peculiarity is the fut. κατεάξω 
Mt 1220 (=Is 42%, but not LXX), which must go with the nouns κατέαγμα 
(BGU ii. 647 bis, P Amh ii. 931°, both 1i/a.p.), ἔαγμα (ap. ἃ. Meyer® 165). 
Possibly -edéas -εάξω -έαγμα may be explained side by side with ἐθελήσας 
ἐθελήσω ἐθελημός, on which see reff. above.2 In LXX (Hab 3! al.) the 
fut. is κατάξω, but κατεάξω in Symm. Ps 475. 


2. Double Augment. 


᾿Αποκαθίστημι (as in LXX)* in Mk 8535 ἀπεκατέστη, Mk 3°=Mt 1215, 
Lk 610 ἀπεκατεστάθη, inserts an augment after both prepositions, which 
seems a well-established vernacular usage. So ἀντεκατέστητε Heb 124 
L* (WH alt.), (παρεσυνεβλήθη Ps 48 (49)! ?! AT). The forms of ἀνοίγω 
described in (1) above are the only surviving exx. of augment attached 
both to verb and preposition. See the list for Attic in Rutherford 
NP 83: of these only ἀνέχεσθαι shows double augment even as a variant 
in NT. Cf, Ac 1815, where 8*B support ἀνεσχόμην, the form attested by 
Moeris as Hellenistic (Ti in loc.).4 From ἀφίημι the impf. is read by 
WH with ἠφ. in Mk 1*4 111°: this form is perhaps not Attic (Meisterh.* 
173, but only one ex., which is not decisive). Note ἐπροεφήτευσεν 
Mk 7° W. 


3. Syllabic Augment for Temporal. 


In verbs which originally began with s or w the primitive syllabic 
augment often leaves its traces behind, contraction following the loss of 
the consonant: thus εἶχον (not ἦχον) for €-exov (-- ἔσεχον), εἵλκυσα for 
é-2dxvoa. In Attic ὠθέω and ὠνέομαι normally augmented éw. (from 
é-Fw.); but ἐξέωσεν Ac 74 S*E (Ti) is the only trace in NT of such 
forms. Ἐργάζομαι and epds. have npy. in Attic,® which prevails in NT ; 


1 See Brngmann-Thumb Gr. 310. 

2 Various unsatisfactory explanations of the word have been given; by 
Thom. Mag. (who would accent κατεάγωσι as perf. subj. act.), by W. (who 
thinks differentiation from fut. of κατάγω adequate—see WM 82), and by 
G. Meyer /.c. (see CR xv. 36). See Cobet on the word (W7' Vatic. lxxix). 

3 Add Letr. 525 (ii/A.D.) and other parallels in WS 8 12. 7 η. But ἀποκ- 
ατεστάθη P Oxy i. 38)? (i/A.D.). See Dieterich 213. 

4° Evoxdéw retains double augment in perf. as late as ii/g.c. (P Amh ii. 37°, 
ἠνώχλησαι). Cf. Reinhold 68. 

5 The perf. ἐώνημαι is found twice in papyri of ii/a.p. (P Oxy ii. 252°, 
P Amh ii. 68°). 

ὁ Perhaps from %-Fepy., Which would account for the different form in the 
perfect. See Meisterhans? 171 ; CR xv. 35. 


190 A GRAMMAR OF NEW TESTAMENT GREEK. [Ἃ8 83 


but the aor. carey. (mid. and pass.) is read by WH with ἐξ four times 
in Paul, B* however thrice opposing. 


4. Dropped Augment. 


In the pluperfect the augment is usually dropped : so Mt 795, Mk 1445 
157-10 [Mk] 169, Lk 1915, Jn 1157, Ac 4533 1498, 1 Jn 219, and temporal 
augment dropped Jn 1180 147. Among these passages only Ac 4°? 
(ἐγεγόνει NAEP al.) shows respectable attestation for the augmented 
form. It appears however unquestionably in Lk 11°? 167°, Jn 952 1144 
(D*om.), Ac 2632 (AL om.). In Attic writers the temporal augment is 
omitted, but not the syllabic, MSS and edd. notwithstanding (see eg. 
Tion Ac 422, Shilleto on Dem. FL p. 38). Attic inscr. down to iii/B.c. 
show 6 augmented forms and no omissions (Meisterh.? 170); and 
Ptolemaic papyri in P Tebt, P Amh and P Fay show 5 augmented against 
2 unaugmented.! The evidence in Schweizer also goes strongly against 
omission. In sucha point the evidence of MSS cannot be trusted far, 
but it may be noted that Jos. (according to Schmidt) prefers omission in 
act., retention in pass. of compound verbs, while in simplicia omission is 
fairly common though not preponderant; even where hiatus is not con- 
cerned. In Polybius the augment greatly preponderates in simplicia, 
though often dropped in compounds, esp. in act. forms.? The pluperf. 
of ἵστημι is a case by itself. The augmentless ἑστήκειν occurs Rev 74 C, 
and is not uncommon outside NT. WH accept throughout the spelling 
ἱστήκειν, in which they think the analogy of the present is to be recog- 
nised, and not mere itacism. WS brings strong arguments against this 
view, which must be regarded as decidedly questionable. As in the case 
of ἴδον below, the complete identity of εἰ and «in popular speech makes 
it only a question of the extent to which the literary tradition was 
remembered. 

Omission of syllabic augment in other tenses occurs only in προορώμην 
Ac 2?5=Ps 15 (16)8 LXX. Here also, since εἶδον arises from é-Fidov, 
would be placed the more than dubious ἴδον in Rev (Ti, WH alt.) and 
ΤΙΧΧ : it seems as though editors and commentators will persist in this 
writing, whatever grammarians say. Those who will may consult 
Gregory’s tabular statement in his Prol. 89. In Jn 5° 10° AL read 
περιπάτει, and in Rom 51% A has ἐλλογᾶτο : such omissions became 
frequent in later times. 

Temporal augment is more often dropped, but only in compounds, 
for we should not accept ὁμοιώθημεν Rom 979 AFGLP=Is 19 (LXX) 
AQ*r. So διερμήνευσεν Lk 3451 διεγείρετο Jn 618 BGL al., ἀνέθη Ac 1678, 
ἀφέθησαν Rom 47=Ps 31 (32)! B, ἀφέθη Mk 13? W, ἀνορθώθη Lk 131% 
ABD al. The first two out of this short list are directly paralleled in 


1Cf. Mayser p. 333 f. 

2 Wackernagel (reviewing Hultsch) Jdg. Fors. v. Anz. 59. 

3 Schweizer 170 and Reinhold 11 may be referred to. If the MSS were any 
evidence on the question of this itacism, we might not unreasonably call in the 
principle of levelling as a vera causa of indic. assimilated to infin. But who 
that has read the papyri would care to build anything on such evidence ? 


§ 83] VERBS. 191 


papyri (CR 1.0.) and in Reinhold’s exx. (p. 66), which casts doubt. on 
Blass’s “clerical error.” Historically ὄφελον is a case of dropped 
augment. 


5. Temporal Augment for Reduplication. 


What has been said of the temporal augment applies to the nearly 
always identical form taken for reduplication by verbs with initial 
vowel. An original distinction is still preserved faintly in the difference 
between the perf. εἴργασμαι and the impf. and aor. npy—see 2 above. 
In ἀφομοιωμένος Heb 78 CDELP, ἀπαλλάχθαι Lk 1258 AA (neither in Ti 
nor WH), the redupl. is dropped.t- WH (App.? 179) would add εὐοδῶται 
1 Co 16? »*B al., which they suggest is a perf. subj. mid., comparing 
those noticed in KBl. § 224. 2 (ii. p. 100): the verb is regularly un- 
augmented in LXX. But the present subj. is exegetically sound (see 
Findlay EGT in loc.), and this perf. subj. type, except for one word from 
Hippocrates,? is entirely confined to verbs where the long vowel is 
radical (kexr@par=kexty-opa, etc.). Nor is that vowel in Attic ever o, 
so that the link for an analogy-process is wanting, and without such a 
process a late denominative verb could never have made such a form. 
Hort favoured this account of εὐοδῶται as fitting in with his view of the 
pres. subj. of -ὅω verbs (App.? 174), on which see below, p. 200. 


6. Initial Diphthongs. 


The short diphthongs are found as follows in the matter of augment 
(reduplication) :—Ai. normally augmented, except ἐπαισχύνθη 2 Ti 116 
NcACDLP al. (WH). The spelling εἰ for ἢ, common in papyri,? is found 
in some MSS of 2 Co 97.--Εἰ, unchanged Gal 2° εἴξαμεν, Mk 154° ἐνείλησεν, 
as in Attic.4—Oi. was augmented ῳ, except κατοίκησεν Mt 41:8 D, ἐνοίκησεν 
2 Ti 15 Do* 17, οἰκοδομήθη Jn 27° NB*TW 33, ἐποικοδόμησεν 1 Co 314 
NAB* al. (ῳ B°C). The omission was Attic only when o preceded a 
vowel. It is common in patr. In οἰκοδομέω WH accept of. in Ac 747, 
and gave it as alt. form everywhere exc. Mt 21%°, Lk 429, Cf. Logiow 
7 (P Oxy i. 1).2—Av. augmented regularly nv. in parts of αὐλέομαι, 
αὐλίζομαι, avédve, occurring 12 times. Once (Ac 1252) εὔξανε in D*, with 
the identically pronounced ev for nu: cf. Blass Pron. 44, who thinks the 
grammarians chiefly responsible for the maintenance of ηυ."---Εὐ. with- 
out augment usually, as in Attic inscrr. since 300 B.c.: so εὐπορέομαι, 
evpopéw, εὐθυδρομέω, εὐνουχίζω, εὐλογέω (with var.). Evdpaivopa, 
εὐκαιρέω, εὐχαριστέω Show nv. in Ac 2° 1721, Ro 171, ed. elsewhere (one 


1 Parallels in papyri in CR 1.0. (4 above). 

2 ἥλκωται (so in K.Bl., but perhaps ἡλκῶται is better for Ionic), which in H 
has βεβρῶται to set the type. 
' 3 (Cf, Meisterh.* 38 f., Blass Pron. 47. 

4 But see Meisterh.? 171, Rutherford VP 244. 

5 Blass notes that w (pron. 6) ‘‘no longer bore much resemblance to οἱ ἢ 
(Ger. 6, pron. nearly like δ). 

8 In G@r.2 8 15. 4 n. 1 he further notes that qv was in later times dissyllabic, 
as seen in the writing yi in NA. 


192 A GRAMMAR OF NEW TESTAMENT GREEK. _ [ὃ 83 


each). In εὔχομαι and προσεύχομαι nv. is general, with var. occasionally 
(ed. twice in Ti, Ac 267° καὶ 27° B*). In the impf. of εὑρίσκω WH read nb. 
Mk 1455, Lk 1948, Ac 711, Heb 11° with alt. each time, but ed. in the other 
tenses.! In εὐδοκέω they read ev. in the Gospels, noting that nv. is some- 
times well supported : in the Epp. nv. 5 times, ed. 6. For words with 
ev- followed by a vowel see 7 below. 


7. Augment and Reduplication in Compounds. 


The primitive rule that in a verb compounded with one or more pre- 
positions the augment or reduplication falls between the last preposition 
and the verb has produced a tendency to place them thus where there is 
no real composition,? and even where the presence of the preposition is 
imaginary. So διηκόνουν (denom. from διάκονος), ἀπεδήμησα (ἀπόδημος), 
ἀπελογούμην (ἀπόλογος) ἐπεθύμησα (*eriOupos), ἐπεχείρησα (ἐπὶ χεῖρα se. 
βάλλω), κατηγόρουν (κατήγορος), συνήργουν (σύνεργος). Sometimes the 
association with the original noun was so vividly present that the rule 
was resisted. Hence ἐπροφήτευσα (προφήτης: Attic mpoed., often in 
LXX, normal in Jos., and as var. in NT, as Jude 14 AC al.), ἐπερίσσευσα 
(περισσός, a deriv. from περί: Phr. mentions ἐπεριέσσευσε (cf. περιέσ- 
σευον Ac 16° E) as a solecism).? In some words the presence of the 
preposition was forgotten (the simple verb being obsolete), and aug- 
ment (redupl.) put at the beginning: ἐκάθευδον, ἐκάθισα, ἐκαθεζόμην, 
ἐκαθήμην, ἠμφιεσμένος (class.), also ἤφιον (above), ἐκάμμυσα (κατ(α)μύω). 
This process became commoner in Byzantine Greek,‘ and survives in 
MGr. The combination of the two tendencies produces the Attic verbs 
with double augment (2 above). The treatment of εὖ as a preposition 
for this purpose—seen in εὐηγγελιζόμην and aor., εὐηρεστηκέναι Heb 115 
NDEP al. (Ti) >—is due to the frequency with which it is used separately 
with verbs, as ed ποιεῖν ete. 


8. Reduplication. 


On analogy of ἕλκω, the denominative ἕλκόω (originally with init. F: 
ef. Lat. wleus reduplicates eiAx. Lk 1650 in all older uncials. In ῥεραντισ- 
μένοι Heb 10°? N*ACD* al. and Rev 1915 (where WH suspect ῥεραμμένον 
and Ti reads περιρεραμμένον with 8*) verbs with init. p have analogical 
redupl.:® so D in Mt 986 ῥεριμμένοι. WH print p (App.? 170) : 7 ancient 


1 Reinhold 65 notes that two MSS which most faithfully represent common 
speech entirely avoid ηὗρον. Ἡὕρισκον once in Hermas (δ) not elsewhere. 

2 See list in K.Bl. § 204 n. 1 (ii. p. 34); also Rutherford VP 79 ff. 

3 K.BI., followed by WS, gives ἐπαρρησιασάμην here, but the old etym. 
from πᾶν is unexceptionable. Blass corrects this in Gr. 39 n.?. 

4 See Hatzidakis 67 f. 

5It is only found when ev- precedes a short vowel: cf. the usage in Att. 
described by Rutherford NP 245. ; 

6 Moeris 459 app. (ed. Koch 417) expressly mentions ῥέρανται as a solecism. 
Ῥεριμμένον Jer 43 (37) A (ἐριμ. B) Jdth 6° A. As early as Homer (Od. 6°”) 
we find ῥερυπωμένα, and in Pindar ῥερῖφθαι (Frag. 314 Bergk). 

7 So Lachm. Lobeck (Par. 14 n.) favours pep. 


§ 83-84] VERBS. 193 


authorities vary—see Κα. Β]. § 67. 3 n. 5, § 200. 1 π. 2. In the latter place 
K.BI. gives classical parallels for this kind of redupl. Verbs in p. usually 
redupl. ἐρρ. (orig. σε-σρ. or Fe-Fp.): so Lk 177, Ac 159, Eph 317, Col 27. 
See G Meyer® 237, also 624. The pp was ultimately made single, to 
resemble other augments: so ἐριμμένοι Mt Uc. NBCL (Ti WH), ἔριπται 
Lk 17? πὲ al. (WH alt.). The substitution of syllabic augment for 
redupl. appears as in Attic in various verbs with init. consonant group. 
So (with var.) ἐμνήστευμαι Lk 127 S*AB*LW, 25 8*AB*C*DW (Ti WH): 
μεμν. Diodorus al. ap. Veitch, and in LXX, on anal. of μέμνημαι. For later 
encroachments of augment on redupl. see Dieterich 214f., CR xv. 36. 
The classical perfect is maintained in two words with init. cons. The 
distinctive εἰ in the perf. εἴργασμαι (=FeFepy.) Jn 371, 1 Pet 4° is practi- 
eally without variant. Cf. Schweizer 170f. WH read ἑόρακα (FeFop.) in 
Paul (1 Co 91 &*B al.,! Col 2! ~*C, 218 NB* al.), and as alt. (with B*), 
7 times in 1 and 3 Jn; but ἑώρακα (-ew) 24 times in Gospels and Ac. 
‘Edpaxa is Attic and original: ἑωρ. (from pluperf. éop.=*nop. 1) is well 
attested in MSS of Attic prose writers. The double reduplication curiously 
known as “ Attic” is maintained intact: so ἀκήκοα, ἀπόλωλα, ἐγήγερμαι, 
ἐλήλακα, ἐλήλυθα, προσενήνοχα. Neither this nor the ordinary redupl. 
is ever dropped, as so often in later times. From λαμβάνω we have only 
the original εἴληφα, εἴλημμαι (=ce-od.). (The similar aor. κατειλήφθη in 
[Jn] 8* seems only a mistake of s: WS wrongly accuses Ti of endorsing 
it.) 
9. Augmented Tenses of Verbs in p. 

As noted in § 41, and for the reduplicated tenses in 8 above, the 
single p prevails over the double. So from ῥαβδίζω 2 Co 11%, ῥαντίζω 
Heb 919-21, ῥαπίζω Mt 26°’, ῥήσσω and cpds. Lk 5° 6458 (but epp. Mt 265°, 
Lk 942), ῥίπτω Mt 157°, Ac 2719, ῥύομαι 2 Co 110, Col 113, 2 Ti 311: 417, 
2 Pet 27 (‘Ti twice epp.). Ἔρρέθην always. 


A, PRESENT STEM. 


(a) THEMATIC. 
ACTIVE VOICE. 


Stem : 
λυο — tyua®/e — &°/e pireo/e σλεξοίξς δηλοῦί. 
loose honour live love sail make clear 


§ 84. Present Indicative— 


Sg. 1 λύω τιμῶ ζῶ φιλῶ πλέω δηλῶ 
2 λύεις τιμᾷς tis φιλεῖς πλεῖς δηλοῖς 
3 λύει τιμᾷ ζῇ φιλεῖ πλεῖ δηλοῖ 
Pl. 1 λύομεν τιμῶμεν ζῶμεν φιλοῦμεν πλέομεν δηλοῦμεν 
2 λύετε τιμᾶτε ζῆτε φιλεῖτε πλεῖτε δηλοῦτε 


3 λύουσι(ν) τιμῶσι(ν) ζῶσι(ν) φιλοῦσι(ν) πλέουσι(ν) δηλοῦσι(ν) 


1In App.2170 ἑόρακα as alternative reading here is presumably ἃ misprint 
for ἑώ. 


194 A GRAMMAR OF NEW TESTAMENT GREEK. _ [ὃ 84 


Imperfect Indicative— 


Sg. 1 ἔλυον ἐτίμων ἔζων ἐφίλουν ἔπλεον ἐδήλουν 
2 ἔλυες ἐτίμας ἔζης ἐφίλεις ἔπλεις ἐδήλους 
8. ἔλυε(ν) ἐτίμα ἔζη ἐφίλει ἔπλει ἐδήλου 

Pl. 1 ἐλύομεν ἐτιμῶμεν ἐζῶμεν ἐφιλοῦμεν ἐπλέομεν ἐδηλοῦμεν 
2 ἐλύετε ἐτιμᾶτε elite ἐφιλεῖτε ἐπλεῖτε ἐδηλοῦτε 
8 ἔλνον ἐτίμων ἔζων ἐφίλουν ἔπλεον ἐδήλουν 


Norr.—1. [π1] sg. impf. the form ἔζην is found in Rom 7° B (where 
33 has ἔζουν, with the mixture noted under 5 below): it occurs in LXX, 
and in all MSS but one of Demosth. Timocr. ἢ. See Mayser 347. The 
analogy of flexions like ἔβη-ν, ἔβης, ἔστην ἔστης accounts for it. 

2. Sporadic instances of the ἔλυα type (weak aor. ending) begin to 
appear in the plural: WS p. 112 cites εἶχαν Mk 87 (NBDA), Ac 28? 
(NAB), Rev 98 (NA), Lk 42 (Ὁ), Ac 81:0 (&), Jn 152224 (D*) ; εἴχαμεν 
2 Jn ® (NA), εἴχατε Jn 941 (“als Var.”—but it does not appear in Ti), 
feyay Jn 1155 (Ὁ) 910 1186 (N*), to which Blass adds Ac 28° Β. 
Scrivener’s list (Codex Bezae p. xlvi) shows that the search has been 
imperfect : he adds from D ἔκραζαν Mt 219, ἔλεγαν Jn 751. 41 8235. 916 1024, 
Lk 23% 24", ὑπέστρεφαν Lk 234°, εἶχαν Mk 81%, Ac 1914, ἔσυραν [which 
however might be aorist] Ac 1419 17%, ἤθελαν Ac 167. Outside Ὁ), 
accordingly, this imperfect is limited to two common verbs, and that 
mainly in. It appears very sparingly before ii/a.p. (see Crénert 210, 
Thackeray 212, Mayser 369, Reinhold 81), but ultimately established 
itself, asin MGr. Mk 87 is the only instance that is at all likely to be 
original. (For -a- forms in flexion of ἥκω, due to its perfect meaning, see 
Prol. 53 and below, § 92.) 

3. Older than this infection is the 3rd pl. impf. in -σαν, which is well 
attested for a non-contract verb in εἴχοσαν Jn 157? ?4 NBL*N* 1, 33 (D* 
εἶχαν): WH App.” 172 note that “in a few other places forms in -οσαν 
[impf. or aor.] have some Western attestation ”—thus Mk 1% ἐφέροσαν, 
616 ἐλέγοσαν (Scrivener).! Instances for the strong aorist are discussed 
below, ὃ 88 (p. 209). Thackeray 213 f. observes that “these forms in -οσαν 
are exceedingly frequent in LXX, being distributed over all the trans- 
lations (except [1-4 K.]) from the Hexateuch to 2 Esdras.” The question 
therefore is how to explain their almost complete absence from NT. 
The extension of the suffix -σαν began in the -w verbs in the earliest 
Greek, and passed into -w verbs in the dialects of Phocis and Delphi (in 
the NW Greek group): see Thumb Dial. 191, Valaori Delphische Dialekt 
60. We may probably regard it as a dialectic form in the Kown, which 
ultimately failed to establish itself. 

4. In Contract Verbs the -cav form becomes rather more prominent, 
though it leaves even less tracein NT. There are about 30 places in NT 
where the 3rd pl. impf. act. of an -ἄω verb is read by one or more of the 
authorities given in MG; but nowhere does Ti cite a form in -ὥσαν. 
From -έω verbs Blass cites ἐθορυβοῦσαν Ac 17° 1), and doubtfully κατοι- 


1 We may add ἐτίθοσαν Ac 8” B, as a clear step towards thematising. 


§ 84] VERBS. 195 


κοῦσαν Ac 24° [D,! with two instances from Hermas. We may add 
διηκονοῦσαν Mk 1611 W. In one place only is there an imperfect from 
the -d verbs, and that is ἐδολιοῦσαν (Rom 31%)? But this is a quotation 
from the LXX, where these forms are common (Thackeray 214).3 See 
Proleg. 52. In NT Greek the -cav form was even ousted from the -μι 
verbs, ἐτίθουν and ἐδίδουν largely replacing ἐτίθεσαν and ἐδίδοσαν. Clearly 
therefore this type was still dialectic, though destined to survive into MGr. 

5. Incipient passage of -dw into -éo forms shows itself here in the 
MSS, as in some other parts of the verb: we may deal with it here in 
advance. Radermacher (p. 73) remarks that it did not extend far till 
ili/A.D., though some traces of the opposite tendency may be found 
earlier. In 3rd plur. impf. we find ἠρώτουν Mt 1528 SBCDX, Mk 41° xC, 
Ac 16°? A, κατεγέλουν Lk 858 D*KX, ἐπετίμουν Lk 1889 AT; but among 
the older uncials only C shows the form frequently (ἐπηρώτουν Mk 938 
105. 10 1218, ἡρώτουν Jn 451, Ac 1%, ἐσιώπουν Mk 934 with N). It is note- 
worthy that C does not show the form in Lk (once in Ac). A few later 
uncials and cursives give -ovy forms in a dozen places. Note the reverse 
change in θεωρῶσιν Mk 5) L, ἐθεώρων Jn 6? A 13, ἐλεᾷ Rom 918 DFG: 
see further p. 196 ff. below. Sometimes the variant -ov has support in 
inferior authorities. Hort and Radermacher are probably right in refus- 
ing to consider the claim of any of these except in Mt 1523, Κοπιοῦσιν 
Mt 658 B 33 is not accepted even by WH: see App.? 173. For other 
instances see p. 197. MGr has taken -éo forms into lst and 3rd plur. 
(and ist sing. mid.) of all -do verbs, and the whole of the impf. : see 
Thumb Handb. § 237 ff. 

6. Verbs in which original F prevented contraction of ew, cov, cot, €o, 
en, are πλέω, δεῖ (impers.), δέομαι, ζέω, πνέω, (ῥέω), (-yéw).4 The NT 
forms are not numerous, but except for πνέει Jn 38 L Chrys and ἐδέετο 
Lk 888 x*C*q) (see below) there is nothing to suggest deviation from 
Attic norm : early papyri likewise preserve this (Mayser 346). Lobeck 
Phryn. 220 ff. collects a good many instances of uncontracted forms from 
later literature, which are more likely to be assimilations of πλέω type 
to Avw than a survival of Ionic, as Phrynichus suggested. 

7. Διψάω and πεινάω have left the -j class, which in Hellenistic only 
retains ζήω and χρήομαι. Πεινᾷ 1 Co 1151: 35 stands without variant. 


Imperative— 
Sing. 2 λῦε τίμα φίλει δήλου 

8. λυέτω τιμάτω φιλείτω δηλούτω 
Plur, 2 λύετε τιμᾶτε φιλεῖτε δηλοῦτε 

3 λυέτωσαν τιμάτωσαν φιλείτωσαν δηλούτωσαν 


Cf. κατωικοῦσαν (ii/B.c.) Magn. 178, ἀξιοῦσαν 474 (Nachmanson 148), The 
accent is proved by Wackernagel ThLZ, 1908, p. 638, 

* Perhaps we should add παρεδιδοῦσαν Ac 164 C. 

° Note that ἐῶσαν is not the only -dw form: there is ἐγεννῶσαν from 
γεννᾶν in Gn 6+, 

* The new present -χύννω probably stands alone in the active, where it avails 
to prevent confusion with fut. xed. Συνέχεον Ac 2157 may be imperf.: see § 95 

VOL. II. PART II.—14 


196 A GRAMMAR OF NEW TESTAMENT GREEK. [Ἃ8 84 


Notre.—1. The imperative of ζήω does not happen to occur. In LXX 
we find ζῆθι (Thackeray 242), ζήτω. 

2. The old Attic 3rd pl. in -όντων, -ώντων, -ούντων, disappeared even 
from Attic in iii/B.c., with an isolated exception : see Meisterhans* 167 f. 

3. Ἑλλόγα Phm 18 s*ACD*FG 33 (cf. Rom 518 and p. 198 below) is an 
instance of -éw form yielding to -άω. So is ἐλεᾶτε Jude 335 NB (see p. 197 
below). 

4, Ἐκχέετε Rev 161 SACP is discussed under χέω in § 95. 


Subjunctive — 
Sing. 1 λύω φιλῶ πλέω and the rest 
2 λύῃς φιλῇς πλέης as pres. indic. 
3 dy φιλῆ πλέῃ 
Plur. 1 λύωμεν φιλῶμεν πλέωμεν 
2 λύητε φιλῆτε πλέητε 
ὃ λύωσι(ν) φιλῶσι(ν) πλέωσι(ν) 


ΝΟΤΕ.--1. The above statement is somewhat doubtfully true about 
the subjunctive of δηλόω : see Proleg. 54. That in 1 Co 1033 παρα- 
ζηλοῦμεν and Gal 417 ζηλοῦτε are subjunctive seems the most probable 
view : it has been accepted by Radermacher 67n., who remarks on the 
assistance given to this fusion at a later period by the convergence of ov 
and ὦ in pronunciation, which he dates in ili/a.D. See below, ὃ 85 (p. 200). 

2. Teva Rom 127° and διψᾷ ἐδ. and Jn 797 (Orig.! διψῇ) continue the 
evidence that these verbs have left the -ἤω class: the LXX of Prov 25?! 
supplies the forms in Rom 1.c. 


Optative— 
Sing. 1 λύοιμι Plur. 1 λύοιμεν 
2 λύοις 2 λύοιτε 
3 λύοι 3 λύοιεν 


Notr.—1. Optatives of Contract Verbs are not quotable in the NT. 
The forms in Hellenistic included according to Moeris (p. 208) ποιῴη, 
after the model of τιμῴη. See further Schmid Atticismus iv. 587f., 
Schweizer Perg. 191. 

2. The paradigm even of the uncontracted verb cannot be completed 
from NT, but its forms are not doubtful. In 3rd pl. we find εὕροισαν 
(strong aor.) in Ac 1727 D*, according to a type common in LXX 
(Thackeray 215). Blass Gr. 46f. thinks this “may be correct, . . . since 
the scribes of D and of its ancestors certainly did not find the optative in 
the living language.” This statement is very questionable, for before the 
date of D there was a curious recrudescence of the optative even in 
illiterate papyri. On the whole it seems better to link D with the 
evidence of the LXX, and regard the -σαν as a dialectic element (here as 
in p. 194, n. 3) which has not touched the NT. Were the phenomenon 
less isolated, it would be tempting to seek light on the history of D. 
See Prol.? 56 ἢ. 


§ 84-85] VERBS. 197 


Infinitive— 
λύειν τιμᾶν ζῆν φιλεῖν πλεῖν δηλοῦν 


Norr.—1. Since -εἰν is historically from -ε(σ)εν, and therefore does 
not contain iota, the contraction from the first shows no ιὲ. Δηλοῦν is 
unchallenged for the period before iv/A.p., with two or three isolated 
exceptions: see the discussion in Prol. 53, and add Thackeray 244. 
Crénert 220 n. gives a number of instances of -οῖν from late MSS. 
Though in five NT occurrences of the infin, B has -oiv thrice, we cannot 
regard this as evidence for the autographs. How the late form arose is 
explained in Prol. Lc. 

2. The printing of « subser. in τιμᾶν and ζῆν is wholly wrong for 
classical texts :1 it never appears in Attic inscriptions—see Meisterhans ? 
175—nor in papyri during the age when the presence or absence of 1 
subscr. counted for anything (Mayser 347). In NT times of course it is 
a mere orthographical question, but there is no reason whatever for 
retaining the «. 

3. Inf. πεινᾶν Phil 4!”. 


Participle. (For declension see § 65.) 


λύ-ων -ουσα -ov τιμ-ῶν -ὥσα -ῶὥν φιλ-ῶν -οὔῦσα -οῦν 
πλέτων 5, 3) fav ” ” δηλῶν ” ” 

Norr.—Passage of -dw into -ἔω occurs in νικοῦντι Rey 217 AC 27 A, 
νικοῦντας 15? C, προσδοκούντων Ac 28° A 25, πλανούντων 1 Jn 27° A; 
ἐλεῶντος Rom 916 SAB*DFGP al. (cf. above, p. 195), and θεωρώντων 
Ac 988 μιὰ afford instances of the converse. 


MIDDLE AND PASSIVE VOICE. 
§ 85. Present Indicative— 


xen? /. use Ser°/. entreat 
Sg. 1 λύομαι τιμῶμαι χρῶμαι φιλοῦμαι δέομαι δηλοῦμαι 
2 λύῃ τιμᾶσαι χρᾶσαι φιλῇ δέῃ δηλοῖ 
3 λύεται τιμᾶται χρᾶται φιλεῖται δεῖται δηλοῦται 


ΡΙ. 1 λυόμεθα τιμώμεθα χρώμεθα φιλούμεθα δεόμεθα δηλούμεθα 
2 λύεσθε τιμᾶσθε χρᾶσθε φιλεῖσθε δεῖσθε δηλοῦσθε 


8 λύονται τιμῶνται χρῶνται φιλοῦνται δέονται δηλοῦνται 


ΝΟΤΕ.---Ἰ. In 2nd sing. the suffix -εἰ of later Attic takes the place of 
τῇ in the word βούλει (LE 2245): the two were no longer equivalent in 
sound as in Attic, for ῃ was identical with ἡ (see p. 97). Since βούλομαι 
was not a “literary” word, as Blass supposed,” we cannot interpret the 
variation by the help of this assumption ; but βούλει may have been 


1The ‘‘high authorities” to whom Hort appeals (Jntrod.? 314) lived un- 
fortunately before the days of scientific philology. 
See Gr. 47. 


198 A GRAMMAR OF NEW TESTAMENT GREEK. [ὃ 85 


stereotyped in general use from use in phrases derived from literature. 
"Own (future) has been levelled (p. 97): its Attic orthography was ὄψει. 

2. In the -dw verbs—and probably in the one -ἤομαι verb—the 2nd 
sing. has established a new analogy form in -oa, drawn from the model 
of the Perfect and from the present of verbs ἴῃ -αμαι. Soin NT καυχᾶσαι, 
ὀδυνᾶσαι, in five places, with no exx. of -@: Blass quotes πλανᾶσαι and 
ἐπισπᾶσαι from Hermas. The other contracta did not follow suit in 
NT. Blass notes αἰτεῖσαι from Hermas 107 &: the -7 form is however 
found in that book, as in Lk 234° φοβῇ (where he suggests that φοβεῖσαι 
for φοβῇ ov would be an easy correction). The LXX has very small 
traces of this formation. Thackeray 218 gives κτᾶσαι Sir 67 and ἀποξεν- 
οὖσαι 3 Καὶ 146 Aquila (ἀπεξ. in the MS) as the only certain exx. from 
Contract Verbs, Moeris contrasts the Attic ἀκροᾷ with Hellenistic 
dxpoaca ; but this is witness no older than the NT, and the same is true 
of Phrynichus. Apart from the solitary form χαριεῖσαι (P Grenf ii. 14 (c)7 
—iii/B.c.), which may be a mere blunder,’ there is accordingly no real 
evidence of this form, outside the LXX, before the second period of the 
Κοινή, which dates roughly from a.p. Wackernagel ThLZ, 1908, p. 639 
thinks it started from the future form πίεσαι, which alone is steadfast in 
LXX: this he derives from the analogy ἵεται : ἵεσαι : : rierac: x. Thence 
naturally φάγεσαι followed, but not in the earliest stratum of LXX :? 
both are firmly established in NT. The future yapieoa: may perhaps be 
accepted in P Oxy ii. 292° (25 a.p.). The extension of the form from the 
-dw verbs to the other contracta may have taken place in i/a.p. Later 
writers show it in abundance (see Hatzidakis p. 188), and it stands in 
M@r now. 

3. Χρήομαι is entered tentatively as assimilated to τιμάω. Only one 
material form occurs in NT, and that is in subj., where the η of the non- 
contracta may have exercised influence. Hermas Vis, 111. 67 has ypaoa 
for 2 sg., and Moeris expressly says χρῆται ᾿Αττικοί, χρᾶται “Ἕλληνες. 
Traces of the old flexion appear in some of its moods sporadically. 

4, Mixture of classes occurs in ἐλλογᾶται Rom 5!% δὲ (accepted by 
WH—rest -εῖται) : the impf. appears in A ἐλλογᾶτο and N* ἐνελογεῖτο. 
Cf. above, p. 196, and see p. 198, π. 8, below. Of a rather different kind is 
ἐμβριμόομαι (simplex once in Xen.), as evidenced by ἐνεβριμοῦντο Mk 145 
NC*W (rest the normal -ὥντο) and pte. (q¢.v.). See below, p. 201. 


Imperfect— 

Sg.1 ἐλυόμην ἐτιμώμην ἐχρώμην ἐφιλούμην ἐδεόμην ἐδηλούμην 
2 ἐλύου ἐτιμῶ ἐχρῶ ἐφιλοῦ ἐδέου ἐδηλοῦ 
8. ἐλύετο ἐτιμᾶτο ἐχρᾶτο ἐφιλεῖτο ἐδεῖτο ἐδηλοῦτο 


1 Grenfell and Hunt thought it ‘due to a confusion of χάρισαι with xapret” : 
they may be right. 

2 Wackernagel says its appearance in A (ter) is to be reckoned among the 
vulgarisms of that MS. 

3 Brugmann Gram.* 348 says, ‘‘The Ionic flexion with a for 7 . . ., which 
from Aristotle’s time passed over into Attic, . . . is explained by assimilation 
to verbs in -aw.” 


$ 85] VERBS. 199 


ΡΙ. 1 ἐλυόμεθα ἐτιμώμεθα ἐχρώμεθα ἐφιλούμεθα ἐδεόμεθα ἐδηλούμεθα 
2 ἐλύεσθε ἐτιμᾶσθε ἐχρᾶσθε ἐφιλεῖσθε ἐδεῖσθε ἐδηλοῦσθε 
8 ἐλύοντο ἐτιμῶντο ἐχρῶντο ἐφιλοῦντο ἐδέονγο ἐδηλοῦντο 


Norr.—1. There is no decisive reason in NT why we should not 
make ἐδεόμην follow ἐλυόμην throughout, and ἐχρώμην take ἢ in 3 sg. 
and 2 pl. Ἐδεῖσθε is absent, and for ἐδεῖτο in Lk 888 BLX 33 ἐδέετο is 
read by N*C*w: AP have the conflate éSeciro. Phrynichus gives the 
uncontracted forms as Ionic in two articles out of three: see Rutherford 
NP 296f. There is good LXX evidence (Thackeray 243) for all three 
types. See Prol. 54 and 234. Late forms like καλέω, there quoted, do 
not invalidate Schweizer’s argument (Perg. 174 n.) that in ἐδέετο et sim. 
we have new analogy forms rather than survivals of Ionic. WH App.? 
173 quote also Jn (3°) πνέει L Chrys}, and inf. πλέειν Ac 27? 112 and 137. 
Their assertion that ἐδεῖτο in Lk L.c. is “better attested” than ἐδέετο may 
only mean that it is in B, or that it is more “correct.” Of the impf. of 
χρήομαι only 3 pl. occurs. A noteworthy form ἐχρήμεθα occurs in 
Gn 267° A (not noted by Thackeray): it is tempting to accept it as 
parallel to ἔζην (above, p. 194), but Brooke and McLean make A the solitary 
witness, and ἐχρησάμεθα matches the surrounding aorists. 

2. No sign appears of the impf. 2 sg. -ᾶσο, corresponding with -ἄσαι 
above. Tenses with -μὴν -co -ro were in much more limited use than 
those in -μαι -σαι -ται, and the force of analogy was therefore much less 
powerful. Grammarians give ἠκροᾶσο (see Hatzidakis 188), but we have 
no reason to believe that it was at all widely used. 

3. For mixture of classes see p. 198, n. 4. 


Imperative — 


Sg. 2 λύου τιμῶ χρῶ φιλοῦ δέου δηλοῦ 

3 λνέσθω τιμιάσθω χράσθω φιλείσθω δείσθω δηλούσθω 
Pl. 2 λύεσθε τιμᾶσθε χρᾶσθε φιλεῖσθε δεῖσθε δηλοῦσθε 

8 λνέσθωσαν τιμάσθωσαν χράσθωσαν φιλείσθωσαν δείσθωσαν δηλούσθωσαν 


ΝΟΤΕ.---Ἰ. The Attic 3 pl. in -σθων has gone the way of the active in 
-OvT@V. 

2. From χρήομαι we have only 2 sg. χρῶ in NT. Xpdo6 can be cited 
from P Oxy vi. 91216 (235 a.D.), vii. 103675 (273 a.D.), and χράσθωσαν 
from P Giss i. 497 (mid. ili/a.D.) and Viereck Sermo Graecus 1689 (81 B.c.). 

3. Νικοῦ Rom 127! A is a case of mixture. Ξυράσθω 1 Co 11° B must 
be aorist like κειράσθω : see below, p. 200, n. 3. 


Subjunctive— 
Sing. 1 λύωμαι τιμῶμαι χρῶμαι φιλῶμαι δηλῶμαι 
2 λύῃ τιμᾷ χρῇ φιλῃ δηλοῖ 
9. λύηται τιμᾶται χρῆται φιλῆται δηλῶται 


Plur. 1 λυώμεθα τιμώεθα χρώμεθα φιλῶμεθα δηλώμεθα 
2 λύησθε τιμᾶσθε χρῆσθε φιλῆσθε δηλῶσθε 
3 λύωνται τιμῶνται χρῶνται φιλῶνται δηλῶνται 


200 A GRAMMAR OF NEW TESTAMENT GREEK. [8 85 


Notr.—l. The contracta have 2 sg. set down in the old form: 
evidence fails for forms in -ca, The solitary form χρῆται in 1 Ti 18 
accounts for the flexion given : see p. 198, n. 3 above. 

2. As in the active, there is strong reason to believe that the -ow 
verbs made subj. identical with indie. pres., at any rate in the plur. : see 
Prol. 54. Hort (WH App.? 174) takes thus φυσιοῦσθε 1 Co 4° and 
διαβεβαιοῦνται 1 Ti 17, as well as the active forms cited above. This is 
certainly true of the former, though it can hardly be admitted that 
Rom 86, an unambiguous conj. deliberat., enforces the same construction 
in Til.c.: the indic. gives as good sense. On the other hand Hort’s view 
(ib. 179), that εὐοδῶται in 1 Co 16? is anything but pres. subj., can safely 
be rejected : see Prol. 54 (also above, p. 191). 


Optative— 
Sing. 1 λυοίμην Plur. 1 λυοίμεθα 
2 λύοιο 2 λύοισθε 
3 λύοιτο 3 λύοιντο 


Norre.—Literature and late papyri,! during the period when the 
optative enjoyed a brief resuscitation, warrant the setting down of the 
old forms. 


Infinitive— 
λύεσθαι τιμᾶσθαι χρᾶσθαι φιλεῖσθαι δεῖσθαι δηλοῦσθαι 


ΝΟΤΕ.---Ἰ, Itacism produces in MSS many spellings with -e, some of 
which raise the question whether the imperative may be read: eg. 
Lk 1417 ἔρχεσθαι all Greek MSS, 1918 πραγματεύσασθαι, Gal 415 ζηλοῦσθε 
NB 33 (where the infin. seems decidedly more probable)—see WH Jntrod.* 
309 f. 

2. The well-attested Hellenistic infin. χρᾶσθαι appears (xaray.) in 
1 Co 918 A 33 Orig.: correct Prol. 54n.8, Early exx. are Syll 1775° 59 
(308 B.c., a rescript of Antigonus), OGIS 214!° (iii/B.c., a dedication by 
Seleucus 1. (?)). But χρῆσθ[αι can be quoted from BGU iv. 1130" (4 B.c., 
Alexandria). In Attica itself χρᾶσθαι prevailed from ii/B.c., though 
instances of χρῆσθαι can be quoted: see Meisterhans* 175. From 
Pergamon Schweizer quotes an ex. of χρῆσθαι from 135 B.c (p. 175). 

3. The accentuation of €vpac6a in 1 Co 116 is questioned. It is 
most generally read évpac@a (“to go shaven”), with the mixed form 
found in the same verse in B—see above, p. 199, n. 3: it is quoted from 
Diodorus. =Zvpé accounts for all the other tenses. In view of the 
association with the aor. κείρασθαι, Heinrici? proposed to read ξύρασθαι, 
aor. of pa, which seems (Lobeck Phryn. 205 n.) to have quite as good 
Hellenistic warrant as ξυράω: ef. Veitch sv. Since ξυράω has no 
probable place in NT, and the change to the present seems without 


1 See Harsing, De Optativi in Chartis Aegyptiis Usu. 
2 Not however in ed.® 


§ 85-86] VERBS. 201 


adequate motive, we may follow WH (App.? 173) in preferring ξύρασθαι, 
which occurs in Plutarch Mor. 336 E and [Lucian] Dea Syra 55 (active in 
Diodorus and Hippocrates). 


Participle— 
λυόμενος -ἡ -ov τιμώμενος χρώμενος φιλούμενος δεόμενος δηλούμενος 


ΝΟοΊΕ.---- Ἐμβριμούμενος Jn 1158. NAU, -opevos BDL αἰ. : see p. 198, 
n. 4, 


(0) UNTHEMATIC. 
ACTIVE VOICE. 


§ 86. [NV.6.—In these obsolescent forms bold type implies 
that the type so printed actually occurs in NT. Forms printed 
otherwise are quotable or inferable from other Hellenistic 
sources. | 

Active forms on these models occur to some extent in 
NT from φημί (1); ἀνίημι, ἀφίημι, παρίημι, συνίημι (2); 
ἀμφιέννυμι, ἀποκτέννυμι, ζώννυμι, ὄμνυμι, σβέννυμι (4). 


(1) ἱστηἃ φη[ἃ (2) THOM] ἴη[ε« (3) διδω[ς (4) δεικ-νῦ,ὺ (5) ἐσ 


Present Indicative— 


Sg. 1 tompe φημί τίθημι δίδωμι δείκνυμι εἰμί 
2 ἵστης τίθης δίδως δείκνυς εἶ 
3 ἵστησι(ν) φησί τίθησι(ν) δίδωσι(ν) δείκνυσι(ν) ἔστι(ν) 

ΡΙ. 1 ἵσταμεν τίθεμεν δίδομεν δείκνύμεν ἐσμέν 
2 ἵστατε τίθετε δίδοτε δείκνῦτε ἐστέ 
3 ἱστᾶσι(ν) φασί (eres διδόασι(ν) δεικνὕασι(ν) εἰσί(ν) 

-ιᾶσι(ν) 

Imperfect— 

Sg. 1 ἵστην ἐτίθην ἐδίδουν ἐδείκνυν ἤμην 
2 ἵστης ἐτίθεις ἐδίδους ἐδείκνυς ἧς, ἦσθα 
3 ἵστη ἔφη ἐτίθει ἐδίδου ἐδείκνυ ἦν 

Pl. 1 ἵσταμεν ἐτίθεμεν ἐδίδομεν ἐδείκνῦμεν ΠΡ 

ἤμεθα 

2 ἵστατε ἐτίθετε ἐδίδοτε ἐδείκνῦύπε ἦτε 
ὃ ἵστασαν ἐτίθεσαν ἐδίδοσαν ἐδείκνῦύσαν ἦσαν 


To these should be added the few surviving forms of the 
mainly literary verb “/, go, viz. 3 pl. pres. -ίασι(ν), 3 sg. impf. 
-yet, 3 pl. τ-ἥεσαν. 


202 A GRAMMAR OF NEW TESTAMENT GREEK. [8 86 


Norr.—1. Forms from the Thematic conjugations invaded these 
Unthematic survivals even in the classical period, as ἐτίθεις -e1, ἐδίδουν 
τους του, and many forms from δεικνύω. In NT none of the models given 
here can be completely evidenced. 

(a) From ἵστημι occur 1 sg. (only Rom 16! in good MSS) and 3 sg. 
pres., but no impf. act. Forms occur from -ἱστάνω and -ordvw: the 
latter is not in LXX, but ultimately secured a permanent place—it is 
MGr. (see Prol.? 55 n.). The impf. καθίστη appears in a fragmentary 
(and rather literary) papyrus of the Roman age, CP Herm 65, From φημί 
we have in NT only the four forms noted above. In the imperf. the 
analogy of ἔστην and ἔβην naturally produced levelling of vowel : hence 
such forms as ἔφημεν in Justin Martyr. In papyri we have some middle 
forms, like φάμενος : their prominence in Herodotus suggests that they 
may have been an Ionic element in the Kowy. How far the old forms 
ἔφαμεν, ἔφατε, ἔφασαν survived in the spoken language is questionable : 
these and other forms not found in NT are omitted above. 

(8) From τίθημι occur 1 sg. and 1 and 3 pl. pres., and from impf. 
3 pL, as well as 3 sg., which is already of the contract type. This was 
extended into 3 pl. ἐτίθουν (Ac 3? 435 817 D*EHLP al., Mk 656 ADNX al.). 
Apparently 76: in Lk 816 D is τιθεῖ, for τίθησι. Mixed forms appear in 
-eridocav B -ετίθεισαν C (Ac 817). The five compounds of in show 
between them 1 and 3 sg. and 2 and 3 plur. pres., but po impf. Non- 
contract forms from -io are common, as ἀφίομεν, impf. ἤφιον. Late 
uncials restore the classical ἀφίεμεν in Mt 6”, Neither in LXX 
(Thackeray 250f.) nor NT are there contract forms, unless we are to 
recognise with WH (so WS 88 14, 16) a type -έω, formed from the future 
-jow in ἀφεῖς Rev 27° and συνεῖτε Mk 817 B*: cf. ἀφῶ in OGIS 201% 
(vi/A.D., rescript of the Nubian king Silko), and ἐπαφῶ in a Lycian inscr. 
But I agree with Thackeray in treating ἀφεῖς as a regular contraction for 
ἀφίεις : see Prol.3 45, where add dvaceis from ἀνασείω, Zauberpap. p. 116. 
In that case συνεῖτε would be aor., which is quite idiomatic, and even a 
plausible reading as differing from συνίετε in v.21; but accidental trans- 
position of letters is more probable. Evidence for -έω is wholly in- 
adequate, and for -éw nil, though the latter is sometimes brought in by 
faulty accentuation (as by Ti in Mt 1318, Rom 911). 

(y) From δίδωμι we find 1,2 and 3 sg. and 3 pl. pres. and 3 sg. and 
pl. impf. But while ἐδίδοσαν survives in Mk 45 C, Jn 19° NB, Ac 164 
NABDE 33 40 61 68 (παρεδιδοῦσαν C—see above, p. 195, n. 4), even the 
risk of confusion with 1 sg. has not kept out the form ἐδίδουν (Mt 13° Ὁ, 
Mk 3° BL, 1573, Jn 193 AD al., Ac 438 (omn.) 16 HLP 27! (A has sg.). 
Contract forms like ἐκδιδοῖ (Arrian Anab. i. 3. 2) do not happen to occur 
in NT, except for διδῶ Rev 39 AC. This might be written δίδω (as MGr.). 
Mk 48 ἐδίδει W follows ἐτίθει : cf. ἐδείδι in BGU 11. 602° (1i/A.D.), and see 
CR xv. 37. 

(8) From classical times forms with -νύω supplanted those in -νυμι 
very largely, especially in active. From NT we can quote only 1 sg. 
δείκνυμι 1 Co 1251] (-vw 33); 3 sg. δείκνυσι Mt 48 (-der δ), Jn 52° (-ver D* 
semel, -υσι semel) ; ἀμφιέννυσι Mt 6°° (Lk 1258 has the vernacular ἀμφιάζει 


§ 86] VERBS. 203 


B or -ἐζει DLT, no doubt from Q'). There is no impf. (ἐζώννυες Jn 21!8, 
ἐστρώννυον Mt 218, Lk 19%*), In LXX ἀπόλλυμι has still some active 
forms (Thackeray 246). 

2. (a) Ἔστι retains its accent at the beginning of a sentence, and 
when it=eaists or ἐδ possible ; also after οὐκ, μή, εἰ, ὡς, Kal, ἀλλά, τοῦτο. 
Only εἶ, among the present forms, is never enclitic. 

(8) Middle forms in the flexion of εἰμί began to come in very early in 
the dialects: cf. Prol. 55f. In MGr εἶμαι εἶσαι ete., they have invaded 
the pres. Class. ἢν 1 sg. has been entirely thrust out by ἤμην (except in 
Ac 2018 D), and ἤμεθα stands side by side with ἦμεν : in Gal 43 both 
appear. The active is some three times as frequent: and ἦσο, ἦτο, have 
not yet begun to show themselves. ‘ 

(y) For ἔστι in practically identical sense occurs in Hellenistic? 
(Paul®, 9451) ἔνι (Ξε ἐν, used in the sense of ἔνεστι or ἔνεισι. In MGr 
this has thrust out ἔστι and εἰσί: the change of each vowel in its new 
form εἶναι (ine for ent) is caused by assimiliation to εἶμαι εἶσαι. 

(δ) Ἦσθα, an old perfect form, was used in Attic (Rutherford NP 
226) for the genuine impf. ἧς of some other dialects. Both survive in 
the Κοινή, but the latter is commoner in NT, where ἦσθα occurs only in 
Mk 1467 (ἧς 1 etc. 13 ete. Kus.) and its parallel in Mt 26°, against seven 
instances of 7s. The reverse was the case in LXX (Thackeray 256). Is 
it possible that this ἦσθα started in Mt under LXX influence, and that 
the text of Mk was harmonised ? 


Imperative— 
Sg. 2 ἵστη τίθει δίδου δείκνυ ἴσθι (ἔσο) 
3 ἱστάτω τιθέτω διδότω δεικνύτω ἔστω OF ἤτω 
Pl. 2 ἵστατε τίθετε δίδοτε δείκνῦτε (ἔστε) 
ε , , / <4 3 + 
3 ἱστάτωσαν τιθέτωσαν διδότωσαν δεικνύτωσαν ἔστωσαν (ἤτωσαν) 


Add -ἴθι from -εἶμι. 


ΝΟΤΕ.---Ἰ. No form from ἵστημι occurs. ἘΕπιτίθει (1 Ti 5633) and 
τιθέτω (1 Co 167), ἀφιέτω (1 Co 71:35) and ἀφίετε συνίετε coincide with 
contract and non-contract τω verbs respectively, and were thus naturally 
preserved. Aidov, διδότω and δίδοτε remain, though in 1 Co 7% A has 
ἀποδιδέτω. Among the -yvp. verbs σβέννυτε 1 Th 5619 stands alone— 
contr. ἀπόλλυε, ὀμνύετε : οἷ. ὑποδίκνυ P Oxy vii. 1066?! (iii/a.D.). 

2. ἜἜστε appears to have become obsolete, or very nearly so. It does 
not occur, in LXX or NT, where γίνεσθε or ἔσεσθε replaces it : see also 
Prol. 180. It can be quoted from Test. Reuben Θ᾽, in one recension, and 
from Eph 5° D°KL al. τοῦτο yap ἔστε γινώσκοντες, as Blass would read 
(p. 320). But ἴστε is overwhelmingly supported, and can be well 
explained as imper. : see Prol.? 245. Blass (p. 308) would make fre in 
1 Co 75 imperative, which would suit very well if any instance of this 


1 Harnack Sayings of Jesus 140, overlooks this certain ex. of the stylistic 
emendation of Mt. See my note in Cambridge Biblical Essays, 486. 
? Late, says Wackernagel Hedlen. 6 n. 


204 A GRAMMAR OF NEW TESTAMENT GREEK. [ὃ 86 


form were quotable. For ἤτω, ἤτωσαν and ἔσο (the last two not in NT, 
ἤτω in Jas 512, 1 Co 1622) see Radermacher Gram. 82, WS 117 n. and CR 
xv, 38, 436: their existence in i/a.D. Hellenistic is not very certain. 
But see Thackeray 256 f. 


Subjunctive— 


Sing. 1 tore 7100 διδῶ δεικνύω ὦ 
a 


2 ἱστῇς τιθῆς διδῷς or διδοῖς δεικνύῃς ἧς 
διδῷ ΟΥ διδοῖ 


and so on like λύω, except for accent: in διδῶ the 7 is 
replaced by o. 


ΝΟΤΕ. 1. Since js and js are only orthographic variants, and ν was 
easily added after a long vowel, the subj. only differed from impf. in 
1 sg. and 1 and 3 pl. Hence in the papyri ἦν, 1.0. 7(v), is very often 
subjunctive : see CR xv. 38, 436, xviii. 108, Prol.* 168, to which a good 
many more instances may be added now. From this start we get ἐὰν 
ἦσθα (Prol.? 1.0.) in LXX and a papyrus of iii/B.c., and ἐὰν ἦσαν P Oxy 
viii. 115715 (iii/a.p.), P Tebt ii. 3331% (iii/a.p.), where a past tense is 
excluded by the context. In NT we find ἦν subj. in Mt 1018 C*, Mk 538 
B*a, Lk 514 ὍΣ, 2028 Ne, 1 Co 164 A, 2 Jn 12 ΝᾺ, 

2. Forms from torn do not occur in NT. From ἀφίημι we have 
ἀφιῇ Mk 11% X, ἀφιῆτε In 16%? L, συνιῶσι Lk 810, Mk 413 (where D¥LW 
1 ete. read συνῶσι), which of course might equally well come from 
(ap)iw, were there any particular reason for so accentuating. Τίθημι only 
gives us 7106 Lk 2043 D and παρατιθῶσι Mk 65] (-θῶσιν AD) 85 (-θῶσιν 
ADNWaA—probably right, the other being assimilated to 641). From 
δίδωμι comes 3 pl. παραδιδῶσι Mt 1019 Cw (leg. aor. or fut.), and 3 sg. 
διδῷ or διδοῖ. These forms remind us of the Hellenistic variations 
between o and described in § 36c; but διδοῖ of course comes easily 
enough from the contract type found in indic. as early as Herodotus. 
The evidence is in no case unanimous: for τῷ we have 1 Co 147 
DcELPw and 1524 SADEP 67%*, for -ot 1 Co 1574 ΒΕ (w read aor.). 
BGU iv. 1127%° (18 B.c.) has προσαποδιδῷ. 


Optative— 

There are no forms in NT, except 3 sg. ein wndecies in 
Lk and Ac, and in Jn 13% SADWI'4AII al. 13 etc. The 
Hellenistic forms of the plural have discarded the primitive 
εἶμεν, εἶτε, εἶεν, in favour of εἴημεν, εἴητε, εἴησαν (as in Ionic), 
due to levelling from the singular: εἶεν has however a better 
record than the Ist and 2nd pl. of the longer form. 


Infinitive— 


ἱστάναι τιθέναι διδόναι δεικνύναι εἶναι 


§ 86-87] VERBS. 205 


Novre.—1. -iorava oceurs in 1 Co 13? SBDEFG 33 (-ἱστάνειν ACK L) 
and in 2 Co 3! FG (-toravew NAC al.). In 2 Co Lec. a contract form 
συνιστᾶν is read by BD* 33, and has a good claim. 

2. ᾿Αφιέναι and τιθέναι have no rivals, nor has διδόναι; nor εἶναι. 
For the -νύναι type may be cited δεικνύναι Mt 167! B (rest -ύειν), ὀμνύναι 
Mk 147! BL unc? (-veew NACWA al.), while -ύειν also occurs in Mt 2674 
(sine var.) : ἀπολλύειν occurs in Rom 1415 FG, 


Participle— 
ἱστάς τιθείς διδούς δεικνύς ὧν 
(like πᾶς) -τ-εῖσα -έν -οῦσα -όν τῦσα τ-ύν οὖσα ὃν 
(8 65. 10) (§ 65. 1) (ὃ 65. 17) like λύων 
Note.—l. Ἐμπίπλημι makes ἐμπιπλῶν, from -dw type. -Iords 1s 


found in 2 Co 43 NCD*FG 33 al. (-avovres A probably, BP 424**— -avrtes 
rest), 64 &*CD*FG 33 (-dvovres BP cu2— -ῶντες rest): WH reasonably 
choose the form -dvovres, which was certainly the prevailing conjugation, 
though -dw forms had a temporary vogue (cf. Thackeray 245), starting 
in old Ionic. 

2. ᾿Ανιέντες Eph 69 and συνιέντος Mt 1319 (DF(L) συνίοντος) συνιείς 
1323 SBD (συνίων ΟἿ ων) stand against συνίων (etc.) in Mk 4° D vt lat 
Rom 811, Neither here nor in indic. do accents in late MSS justify our 
bringing in a type -.é, which never shows itself in distinct contracted 
forms. Τιθείς stands in Mk 10° (τιθῶν 1 28 13-346), Lk 4*° BD vg, 
Judet : cf. ὑποτιθοῦσα BGU i. 350 (ii/A.D.), ἐπιτιθο(ῦντι) P Oxy vi. 986 
(i/a.D.), and LXX once (1 Es 480 BA) ἐπιτιθοῦσαν. 

3. Διδούς naturally prevails, since except for nom. sg. m. its flexion is 
identical with that of λύων. But ἀποδιδοῦν Rev 22? A al. (a correction 
to secure concord), mapadidav Mt 264° x*, Mk 1445 D: διδώς in Jn 6%° D 
is a case of virtual itacism, like ἐδίδων 3 pl. impf. in Mk 157% Μὰ, 
Thackeray 250 gives διδοῦντι from Pr 268 δὲ (-όντι BA). Cf. διδοῦντος 
BGU i. 8672, ἀνδιδοῦντα ab. 44, P Oxy iii. 53211- 81] ii/a.p. 

4, ᾿Απόλλυμι, as in its whole active flexion, makes a thematic pte. 
(Rev 914), but -vs is found in other verbs. Thus ὑποζωννύντες Ac 2717, 
δεικνύντος Rev 228 δὲ cul! (-vovros A 046 al.), ἐπιδεικνύς Ac 1878, ἀποδεικ- 
vuvta 2 Th 2* (-vovra AFG). In Rev the -vwv type is ὦ priori more 
probable. 


MIDDLE AND PASSIVE VOICES. 
§ 87. Present Indicative— 


no sit κει lie 

Sing. 1 ἵσταμαι τίθεμαι δίδομαι δείκνῦμαι κάθημαι κεῖμαι 
2 ἵστασαι τίθεσαι δίδοσαι δείκνυσαι κάθῃ κεῖσαι 

8 ἵσταται τίθεται δίδοται δείκνυται κάθηται κεῖται 

Plur. 1 ἱστάμεθα τιθέμεθα διδόμεθα δεικνύμεθα καθήμεθα κείμεθα 
2 ἵστασθε τίθεσθε δίδοσθε δείκνυσθε κάθησθε κεῖσθε 


8 ἵστανται τίθενται δίδονται δείκνυνται κάθηνται κεῖνται 


1 Προσδιδέναι BGU iv. 11154 (18 8.0.). 


206 A GBAMMAR OF NEW TESTAMENT GREEK. [8 87 


Imperfect Indicative— 


Sing. 1 ἱστάμην ἐτιθέμην ἐδιδόμην ἐδεικνύμην ἐκαθήμην ἐκείμην 
2 ἵστασο ἐτίθεσο ἐδίδοσο ἐδείκνυσο ἐκάθησο ἔκεισο 
8 ἵστατο ἐτίθετο ἐδίδοτο ἐδείκνυτο ἐκάθητο ἔκειτο 
Τί». 1 ἱστάμεθα ἐτιθέμεθα ἐδιδόμεθα ἐδεικνύμεθα ἐκαθήμεθα ἐκείμεθα 
ῶ ἵστασθε ἐτίθεσθε ἐδίδοσθε ἐδείκνυσθε ἐκάθησθε ἔκεισθε 
8 ἵσταντο ἐτίθεντο ἐδίδονγο ἐδείκνυντο ἐκάθηντο ἔκειντο 


Notr.—1. As in LXX and Ptolemaic papyri (Thackeray 245, 
Mayser 351f.) the unthematic forms are much better preserved in the 
Middle than in the Active. Additional verbs of these classes, besides 
some of those named under the Active, are (1) δύναμαι, ὀνίναμαι, κρέμαμαι, 
ἐπίσταμαι, πίμπραμαι, (4) μίγνυμαι, ῥήγνυμαι, ἀπόλλυμαι. 

2. Two roots, peculiar in that they appear in the Middle (in Sanskrit 
as well as Greek) without vowel-reduction, have ἃ partial flexion in this 
class, preserved mainly by virtue of their similarity to the Perfect. Κεῖμαι 
is common and presents no irregularities : 2 sg. and pl. pres. and 1 and 2 
sg. and pl. impf. do not happen to occur in NT. From the root jo only 
κάθημαι has survived, and augment and accent alike show that it was not 
felt to be a compound, The impf. only occurs in 3 sg. (wndecies) and 
3 pl. (Mk 3°? Aal.). In LXX the regular 2 sg. κάθησαι still occurs, but 
κάθῃ (from κάθομαι---οἵ. imper.) appears without variant in Ac 23%, the 
only NT occurrence : cf. P Oxy i. 33 18 (ii/a.p.). 

3. The two tenses can be made up for the ἵσταμαι type out of the 
four verbs that contribute forms, except for the 2 sg. impf., the form of 
which is not quite certain: Dt 28°, 15 48° have ἠπίστω preserving an 
old classical alternative (Kiihner® 1. ii. § 213, n. 1), which Moeris even calls 
Attic against Hellenistic -aco (Schweizer Perg. 168, who quotes διίστω 
from Pergamum). δΔύνασαι is normal, appearing six times without 
variant : δύνῃ, from dvvopa,! occurs in Mk 952 SBDLWA (-aca: ACNX), 
23 S*BDNWA (-aoa N°CACLX), Lk 16? NBDPW (al. future), Mk 14° B 
(-aca. SACDLal.), Rev 25, Similarly for (ἐξ) -ἐκρέματο Lk 1945 ADLQRW 
we have ἐξεκρέμετο NB. 

4, From ddiewa we have 3 sg. and 3 pl. pres. ᾿Αφίενται is in Mt 925 
NB, Mk 2° B 28 33, 29 SB 28 565, Jn 2038 W and later uncials, Lk 747 W ; 
while we find ddiovra Jn 2078 B*, Mk 2° A, Lk 747 F. Τίθεμαι shows 
1 sg. and 2 pl. pres., 3 sg. and pl. impf. Προσετίθοντο is read by cu.‘ 
(incl. 1) in Ac 5!*. Cf. παρακατατίθομαι BGU i. 326 (11/a.D.). 

5, Δίδομαι has 3 sg. and 1 pl. pres., but in impf. διεδίδετο Ac 488 
NB*ADE, παρεδίδετο 1 Co 11° SB*ACDEFGK 33, with P and late 
authorities for -oro. See under the Aorist, and cf. LXX in Thackeray 250. 

6. For -νυμαι forms may be quoted ἐνδείκνυνται Ro 215, ῥήγνυνται 
Mt 917, διερήγνυτο Lk ὅδ A ὑπο], ἀπόλλυμαι Lk 1517, -υται 1 Co 81! 
NABDP 33, Mk 222 BL, -ύμεθα Mt 875, Mk 438, Lk 824, -uvra Mt 917 
NB 1 18, ἀπώλλυντο 1 Co 10° NBA, 10! A, σβέννυται Mk 948, -υνται 
Mt 258. The are no -vo forms at all. 


1 WS 118 wrongly calls it contracted. The type δύνομαι occurs fairly often 
in papyri: see Thackeray 249, Mayser 355, CR xviii. 112. 


§ 87] VERBS. 207 


Imperative— 
Sing. 2 ἵστασο τίθεσο δίδοσο δείκνυσο 
3 ἱστάσθω τιθέσθω διδόσθω δεικνύσθω 
Plur. 2 ἵστασθε τίθεσθε δίδοσθε δείκνυσθε 
3 ἱστάσθωσαν τιθέσθωσαν διδόσθωσαν δεικνύσθωσαν 


ΝΌΤΕ.---Ἰ. The imper. κάθησο, still found in LXX, is supplanted in 
NT by κάθου (as from κάθομαι), which occurs six times with no trace of 
the older form. No other imper. appears; nor any imper. of κεῖμαι. 
Καθήσθω etc., and κεῖσο etc., may be postulated as the only conceivable 
forms if the tense was ever wanted. 

2. -ioraco (mepi- bis, ap- semel in late MSS) has no variant form. 
(See p. 206, n. 3.) Παρατιθέσθωσαν 1 P 419 is the only quotable part of 
τίθεσο ; while συναναμίγνυσθε 2 Th 831: Ew (an itacism) alone represents 
the other types. “Emurifov (as from a contract verb) occurs in 1 Ti 52? 
D: we might equally well write ἐπιτίθου, as from -τίθομαι. 


Subjunctive— 
Sing. 1 δύνωμαι τιθῶμαι διδῶμαι δεικνύωμαι καθῶμαι 
2 δύνῃ τιθῆ διδῷ δεικνύῃ καθῇ 


and so on like λύωμαι, except for accent: in διδῶμαι the ἡ 
is replaced by ὦ. 


Norr.—Kadjo8e Lk 2259 B*TA represents the subj. of κάθημαι. 
From the rest no forms occur except δύνηται δύνωνται. 


Optative— 
Sing. 1 δυναίμην Plur. 1 δυναίμεθα 
2 δύναιο 2 δύναισθε 
ὃ δύναιτο 3 δύναιντο 


Norr.—The only optative from which forms occur (1 sg. and 3 pl.) 
is cited by itself, as there is no evidence that NT writers would have 
used any of the rest. 


Infinitive— 
ἵστασθαι τίθεσθαι δίδοσθαι δείκνυσθαι καθῆσθαι κεῖσθαι 


ΝΟΤΕ.---ΑἸ] these types occur without alternatives. This fact makes 
it very improbable that we should accent πιμπρᾶσθαι or ἐμπιπρᾶσθαι in 
Ac 28° as if from a contract verb, which we should naturally accept 
in the active. 


Participle— 


ἱστάμενος τιθέμενος διδόμενος δεικνύμενος καθήμενος κείμενος 


Norr.—All these are well represented in NT, and there are no 
alternative forms. 


208 A GRAMMAR OF NEW TESTAMENT GREEK. [8 88 


B. STRONG AORIST STEM. 


§ 88. On the formation of this stem see above, § 82. 
It only concerns non-contract -o verbs and the verbs in -μι. 
There is no present tense for this stem. 

(a) Thematic, (Ὁ) Unthematic. 

(a) Bars/z cast; (Ὁ) (1) στή («) stand, (2) θη. place, (3) 6°/, 
give, (4) dv sink. 

(Ὁ) Like (1) are ἔβην from βαίνω and (opt. only) ὠνάμην 
from ὀνίνημι. 

Like (2) are compounds of ime. 

Like (3) is ἔγνων from γινώσκω, but varying strongly in 
parts. 

In (4) ἔδυν is nearly obsolete, and ἔφυν (from φύω) 
seems wholly so. 


ACTIVE VOICE. 


Indicative— 
Sing. 1 ἔβαλον ἔστην sing. supplied by ἔγνων ἔδυν 
2 ἔβαλες ἔστης Ist aor., ἔθηκα, ἔγνως ἔδυς 
ὃ ἔβαλε(ν) ἔστη ἔδωκα ἔγνω ἔδυ 
Plur. 1 ἐβάλομεν ἔστημεν ἔθεμεν ἔδομεν ἔγνωμεν ἔδυμεν 
2 ἐβάλετε ἔστητε ere ἔδοτε ἔγνωτε ἐἔδυτε 


3 ἔβαλον ἔστησαν ἔθεσαν ἔδοσαν ἔγνωσαν ἔδυσαν 


Norr.—1l. In Thematic Verbs there is a large infusion of Weak 
Aorist terminations, as in other parts of the system. It began in Attic, 
where εἶπον is inflected -ον -as -e(v), τατον -άτην, -ouev -are -ov: similarly 
ἤνεγκον (from φέρω), but with 1 pl. -αμεν (Rutherford NP 219f.). In 
these verbs the double tenseformation was primitive: see Brugmann Gr. 
322. Cf. Prol. 51,1 and Thackeray 210 f., who shows that except in these 
two verbs the -a forms did not become common till i/a.p., when the 
papyri begin to show them freely: cf. CR xv. 36, xviii. 110. In NT 
these two verbs are almost exclusively found with -a- in indic.: εἶπες 
Mk 1982 SDL al., Jn 417 SB*, with εἶπον 1 sg. usually, are about the 
only exceptions, with εἶπον 3 pl. rarely. Ἔπεσα -are -av were 
helped towards predominance by the o; ἤλθαμεν -αν and once -a 
(Rev 10%) are found ; εἶδαμεν ταν and less certainly -are, with εἶδα in 
Rev ; εὕραμεν and -ay, but only -ov in 1 sg. ; -eiAare -αν : in other verbs 
the instances are more isolated—é8aday (ἐπ-) is thrice provisionally 
accepted by WH, with ἐξεβάλαμεν rejected in Mt 732, ἐλάβαμεν -are -αν in 
Lk 55,1 Jn 357, Jn 113, ἔπιαν in 1 Co 104, ἀπέθαναν in Mt 832, Lk 2031, 
Jn 853, See WH App.? 171f, WS 111f., and for exx. in D see 


1 Correct the total given there for verbs forming strong aor. act. or mid. : 
there are over forty. 


§ 88] VERBS. 209 


Scrivener Codex Bezae xlvi. The increasing prominence of these forms, 
especially in the plural, is thus well seen. The MGr aorist flexion -a -es 
(§ 89) -ε -ape -ere (and -are) -av, shows how these forms lived on. Cf. on 
impf. above, p. 194, n. 2. 

2. In 3 pl. the suffix -οσαν---88 in the impf. (p. 194, n. ὃ, 4)—obtained 
a footing for a time, but it makes no claim in the NT for the aorist except 
in παρελάβοσαν ἃ Th 3° x*AD* 33. BGg read παρελάβετε, which WH 
put in their text, remarking that the uniqueness of this termination in 
Paul renders it “somewhat suspicious”: the mistake may have arisen 
from an ocular confusion if Trapadocin stood in the line above just 
over tTrapeAaBete. Blass 46 argues that this and the impf. are prob- 
ably authentic, since they could not have been very familiar to the 
scribes except in contract verbs. Scrivener cites from D ἤλθοσαν Mk 8:1 
988 εἴδοσαν 9°. As observed above, the form had only a temporary vogue, 
except in the Contracta (impf.), where it remains in MGr. 

3. The aor. indic. of -ἴημι is supplied wholly by ἀφῆκα, as is that of 
τίθημι by ἔθηκα, and of δίδωμι by ἔδωκα, with the significant exception of 
παρέδοσαν in Luke’s literary Preface (17): contrast the Middle below. 
In Mk 718 W reads παρέδοτε, whence ἔδοτε is marked as NT above. 
Ἔστην and -έβην between them form a complete flexion (exe. 2 sg.), and 
ἔγνων is complete. Εδυν occurs in 3 sg. Mk 1°? RACLWTIATI al. (ἔδυσεν 
BD 28), and in 8 pl. παρεισέδυσαν Jude * SACP ete. (-εδύησαν B alone). 
If we read ἔδυσεν in Mk /.c., both passages show the weak aorist dropping 
the transitive sense, which is likely enough. The aor. pass in Jude 4 is 
parallel to ἐφύην which has supplanted ἔφυν, and it is quoted by Veitch 
from Hippocrates and Hesiod (ἢ). Whichever reading be adopted, the 
solitary survival of ἔδυ in Mk l.c. seems highly improbable. 


Imperative— 
Sg. 2 Bade στῆθι) -βηθι θές ἄφες δός γνῶθι 
-στα { -βα 
3 βαλέτω στήτω -βάτω θέτω ἄφέτω δότω γνώτω 
Pl. 2 βάλετε στῆτε -βατε θέτε ἄφετε δότε γνῶτε 


8 βαλέτωσαν στήτωσαν -βάτωσαν θέτωσαν ἀφέτωσαν δότωσαν γνώτωσαν 


Norr.—1. From εἶπον the imper. has exclusively weak aor. endings, 
except that εἰπέ also occurs: εἰπόν 1 is said by WH (App.? 171) to stand 
chiefly before consonants. They accept -éveyxe four times, -ένεγκον 
once (Mt 82), and ἐνέγκατε without variant. From ἔπεσον however -πέσε 
and πέσετε are best attested ; from ἦλθον ἔλθατε ἐλθάτω, though B has -e- 
five times. 

2. In compounds -στηθι and -στᾶ, -βηθι and -8a alternate without very 
clear rationale. The short forms are found in Attic poets (Blass? 50 n.) 


1 The imper. εἰπέ, ἐλθέ, εὑρέ were oxytone in Attic and the Κοινή, as were 
ἰδέ and λαβέ in Attic : see Kiihner® § 217. 3. ἃ (1. ii. 84). It is a survival of the 
original accent: cf. Brugmaun Gr.4 183, who accepts mé, φαγέ as well. Blass 
45 follows Lobeck Phryn. 348 in accenting the imper. εἶπον. 


210 A GRAMMAR OF NEW TESTAMENT GREEK [ὃ 88 


but v. inf.—and in MGr: so Hatzidakis 101, where ἀνέβα ἀνεβᾶτε, ἔμβα 
ἐμβᾶτε and the like are cited. But in some MGr dialects ἀνεβῆτε ete. 
appear (zb.). This suits the fact that the ἡ forms sg. and pl. still survived 
in ancient Kow7: in LXX exclusively in -βαίνω, though -ora occurs as 
well as τστηθι.1 In NT στῆθι occurs thrice simple and eight times com- 
pound ; -ora occurs Ac 127, Eph 5'*, Ac 911 B, 1050 D*, 1119 D* (against 
ἀναστάς). For -Ba we have Rev 41 (exe..A), Mk 15°? L, Mt 1730 XB 1 
etc. 13 etc. (against -βηθι CDw) ; for -βηθι also Mt 274°, Lk 141° 195, Jn 449 
73, Ac 100, The other persons are always in ἡ with στήτω ete. ; but the 
analogy of τίμα τιμᾶτε has made καταβάτω five times (only Mk 131° XA 
-βήτω) and dvdBare Rev 1115 SACP (only 046 -8yre).? The accent of 
τίμα τιμᾶτε combines with ΜΟῚ ἀνέβα aveBare (see above) to make it 
probable that we should write ἀναβᾶτε in Rev 1115, instead of following 
the older ἀνάβητε. It may be noted that ἀνάβα appears on an Attic 
vase-painting: see Kretschmer Vaseninschriften p. 197. Its occurrence 
in Attic comedy (as Arist. Ran. 35, ἔμβα 2b. 377) suggests that it was Attic 
vernacular already, and not really poetical, though used by Euripides. 

3. “Ades ἄφετε became a quasi-auxiliary: see Prol. 175. és θέτε, δός 
δότω δότε, γνῶθι γνώτω γνῶτε, show no signs of obsolescence. The 3rd 
pl. does not happen to occur. MGr θές, δός and ἄς (-- ἄφες) show that 
the 2 sg. was the most firmly rooted. 


Subjunctive— 
Sg. 1 βάλω στῶ θῶ ἀφῶ δῶ γνῶ 
2 βάλης στῇς θῇς ἀφῇς δῶς, Sots γνῷς 
3 βάλῃ στῇ θῇ ἀφῇ δῷ, δοῖ, δώῃ γνῷ, γνοῖ 
Pl. 1 βάλωμεν στῶμεν θῶμεν ἀφῶμεν δῶμεν γνῶμεν 
2 βάλητε στῆτε θῆτε ἀφῆτε δῶτε γνῶτε 
8 βάλωσι στῶσι(ν) θῶσι(ν) ἀφῶσι(ν) δῶσι(ν) γνῶσι(ν) 


ΝΟΤΕ.--1. Στῶ suffers somewhat, like the rest of the paradigm of 
ἔστην, from the competition of the weak aor. pass., but it occurs (3 sg. 
and pl., and 2 pl.) six or seven times. -B@ is found in 3 sg. From 
τίθημι and -ἴημι forms are common, and abundant from the two -w- verbs. 

2. Where 6 occurs in the flexion of δῶ and γνῶ there was a strong 
Hellenistic bias towards οἵ: in papyri it affects 2 sg. as well as 3 sg.,8 
and D shows this in Lk 125°. The analogy of contract verbs might work 
directly on an aor. form—cf. ἀναβᾶτε just discussed—and there would be 
a natural tendency to eject unique subj. forms like -és and -o. But it 
may be also observed that there are several instances of Hellenistic οἱ 
replacing earlier @: see CR xv. 37, 485. The figures for NT MSS are 
interesting. NS has dot $ times, B τὺ, ἃ 8, C 3, D&, L 9, W ὁ in the 


1 Thackeray 254: he says -ora is poetical in LXX. 

2 [Against this we have ἀνάβητε without variant in Jn 7°.—Ep.] 

3 See Prol.2 55 π.3 and add—e.g. Mél. Nic. 185 (Ptol.), P Tebt ii. 409° 
(ἀν. δ) 


§ 88] . VERBS. 911 


Gospels: in Paul δὲ 34, B4,A ὃ, Ο 9,7, 1, 6 5 ΚΤΡ 2. (These last only 
stand for = in Paul we have to add the places where δώῃ appears, viz. 


δ 2, B 0, A 2,C 1, Ὁ, ὅ, α 8 and KLP 2.) Thus the 5-text shows δοῖ 
most, importing it even into Lk (22* D): otherwise Luke and Paul show 
no trace of it (exc. in 1 Th 5!5, where & joins DG). An obviously ver- 
nacular form—as its papyrus record shows—it may safely be assumed 
right in Mk 4539 NBD, 1410} BDW, 857 XB, Jn 13? SBD and perhaps 
Jn 1322 D, Though a late form of the opt. coincides with it, there is 
not the slightest syntactical reason for doubt that in NT it is always 
subj., as W. F. Moulton proved long ago (WM 360n.). With dot goes 
yvoi, read by WH in Mk and Lk (Mk 5*? ABDLW, 93° SBCDL, Lk 1915 
NBDL 33) against γνῷ in Jn 7°! 1157 (yvot D*) 1431, and asv.l. in Mk and 
Lk Lc. (A bts, δὲ and C semel, W bis). 

3. A third form, δώῃ, occurs in Paul: δὲ 2, B 9, A 2, C 4, Ὁ, ὅ, G 8, 
and even KLP 2—cf. Jn 1516 in 33 and late uncials. For proof that the 
form dan is subj. as well as opt.—the different placing of ¢ swbscr. is only 
orthographical—see Prol.? 55 and 193 f.: a clear instance of γνώῃ subj. 
is there cited from Clement, with a ref. to Reinhold 90f. for δώῃ in 
apocrypha, and to a new reading (ἀποδούηι) in a pre-Christian papyrus.! 


Optative— 


The 3 sg. δῴη occurs in Paul ὁ (Rom 155, 2 Th 316, 2 ΤΊ 116-18), and 
in late texts of 2 Ti 27 414, Aot was also a Hellenistic form. From 
thematic verbs the flexion was βάλοιμι -οις -οἱ, -οιμεν -oure -ovev with no 
possible alternatives except in 3 pl. Here the form in -σαν appears in 
εὕροισαν Ac 1727 D*, as in LXX: see above, § 84, p. 196. 


Infinitive— 
Βαλεῖν στῆναι -βῆναι θεῖναι ἀφεῖναι δοῦναι γνῶναι 


Norr.—l. All these are well represented, and there are no signs of 
variants (such as orava, -βάναι, δῶναι, yvovva, due to mixture with 
present forms, and confusion between the two -ω- roots), which appear in 
papyri: see CR xv. 37, 435. 

2. ᾿Ανενέγκαι 1 Pet 2° is the only weak aorist form. 


Participle— 
βαλών στάς βάς θείς ἀφείς δούς γνούς 
-οῦὖσα -όν -ἄσα -ἀν -εἶσα -€v -οὔσα -όν 


ΝΟΤΕ.--Εἴπας appears, though rarer than εἰπών: the two occur 
together in Jn 1158 BC*. WH reject the oblique cases. 
SS EEE Eee 

1[Mr. H. Scott cites Test. Simeon 4° ἵνα δῷ ὁ θεὸς χάριν, v.12. δῴη, δώει. 
Test. Reuben 4! ἕως οὗ ὁ κύριος δώῃ ὑμῖν cbGvyov.—ED. ] 

VOL. II. PART. 11:-- οὖς 


212 A GRAMMAR OF NEW TESTAMENT GREEK. [8 88 


MIDDLE VOICE. 


Indicative— 
ἐβαλόμην ἐθέμην ἐδόμην 
ἐβάλου ἔθου ἔδου 
ἐβάλετο ἔθετο ἔδοτο 
ἐβαλόμεθα ἐθέμεθα ἐδόμεθα 
ἐβάλεσθε ἔθεσθε ἔδοσθε 
ἐβάλοντο ἔθεντο ἔδοντο 


ΝΌΤΕ.---Ἰ. Forms of the thematic verb with a include ἀπειπάμεθα 
2 Co 45, and (as in act.) the quotable forms of the aor. of αἱροῦμαι : 
ἐξειλάμην, εἵλατο and ἀν- and ἐξ- εἰλατο. Inthis word the resemblance to 
the weak aorist of a liquid verb no doubt helped the mixture: see under 
the Participle. 

2. Mk 8: B has ἐπελάθεντο, which occurs 8 times in LXX, according 
to Thackeray, who calls in the analogy of ἐτίθεντο, and the occasional 
3 pl. -ecav for -οσαν. That the variation affects this verb only, and in 
LXX as well as NT, makes some special cause probable. 

3. ᾿Ἐξέθοντο Ac 187° D shows thematising, also συνεπείθοντο Ac 24° 
33 and 237° συνέθοντο H*. 

4. For thematic edero there is quotable ἀπέδετο Heb 1216 AC, ἐξέδετο 
Mt 2135 S*B*CL, with A added in the || Lk 209, and AK in Mk 121. In 
Heb l.c. the “correct” form is read by NDKLP 33, and is what we 
expect: in Mk 12! the vernacular flexion is equally to be expected, and 
we find Mt and Lk faithfully copying it. W has ἐξέδοτο only. None of 
the other moods of ἐδόμην occur in NT, 


Imperative— 
βαλοῦ θοῦ 
βαλέσθω θέσθω 
βάλεσθε θέσθε 
βαλέσθωσαν θέσθωσαν 


Nore.—The three occurrences of -θοῦ and θέσθε are in Luke and 
Paul. 


Subjunctive— 
βάλωμαι θῶμαι 
βάλῃ θῇ 
βάληται θῆται 
βαλώμεθα θώμεθα 
βάλησθε θῆσθε 
βάλωνται θῶνται 


ΝΟΤΕ.--Α ποθώμεθα in Rom 1313 is the only ocourrence. 


§ 88] VERBS. 218 


Optative— 


In the NT no forms occur from unthematic verbs except ὀναίμην 
(Phm 39) from ὀνίναμαι: cf. ὄναιντο in Audollent 92° (iii/B.c.). Γένοιτο 
from γίνομαι is also a living form in the vernacular, esp. in the expres- 
sion μὴ γένοιτο. Λάβοιντο can be quoted from a very illiterate Christian 
letter of iv/v A.D., P Giss 412, The forms are the old ones throughout, 
so far as they survive at all. 


Infinitive— 
βαλέσθαι θέσθαι 


NotE.—EvpacOa can be cited from P Oxy ix 190418 (299 a.p.). 
There are no NT parallels: see below. 


Participle— 
βαλόμενος θέμενος 
Notre.—Etpdpevos (Heb 913) is a well-attested form, whose passage 
into the weak flexion is explained, like εἵλατο etc. above, by the close- 
ness of the weak aor. in liquid verbs. So γενάμενος, which is plentiful 
in papyri, but very rare in good NT texts (¢g. Lk 22** δὲ͵ 24%? B): see 
Proleg. 51 n.4 


PASSIVE VOICE. 


Indicative— 
ἠλλάγην ἠλλάγημεν 
ἠλλάγης ἠλλάγητε 
ἠλλάγη ἠλλάγησαν 
Imperative— 
ἀλλάγηθι ἀλλάγητε 
ἀλλαγήτω ἀλλαγήτωσαν 
Subjunctive— 
ἀλλαγῶ ἀλλαγῶμεν 
ἀλλαγῇς ἀλλαγῆτε 
ἀλλαγῇ ἀλλαγῶσι(ν) 
Optative— 


Not found in NT: it runs ἀλλαγ-είην -είης -είη -είημεν 
-εἰητε -εἰησαν. 


Infinitive— 
ἀλλαγῆναι 


214 A GRAMMAR OF NEW TESTAMENT GREEK. [§ 88-89 


Participle— 
ἀλλαγείς -εἶσα -έν 


Norr.—Twenty-five roots are found in NT with this strong aorist, 
and some of them belong to the post-classical age, so that the formation 
was still alive. Sometimes it even ejected an older weak aorist: 6.6. 
ἠγγέλην for ἠγγέλθην---ἃ denominative like ἀγγέλλω could not make 
strong tense-stems in earlier Greek. 


C. WEAK AORIST STEM. 


§ 89. For the general formation of this stem see above, § 82. 
It proved in later Greek a pivot of the verb, very frequently 
producing new present stems. See Thumb Handd, 143 f. 


Norr.—1l. The development in the MGr verb throws much light on 
tendencies already visible in NT Greek. Strong and weak aorists are 
now fused, and the characteristic a of the weak aorist endings 
dominates the active, banishing completely the endings with -o-. The 
impf. has taken the same set of endings—-a -es -«, ταμε -ere (and -are) 
-av(e). Accordingly the Hellenistic tendency to assimilate the two 
aorists, and at the same time to keep the imperfect in touch with the 
aorist, has worked itself out to asymmetrical result. The same historical 
connexion is seen in the stems. Thumb (l.c.) remarks that the MGr aor. 
act. “corresponds exactly” to its old Greek predecessor. “Only in a few 
cases the sigmatic aorist has encroached upon the territory of the non- 
sigmatic ; thus ἐκέρδεσα fr. κερδαίνω, ἐσύναξα fr. συνάγω (συνάζω), ἐπρό- 
σεξα fr. προσέχω, ἁμάρτησα fr. ἁμαρτάνω." ΑἸ] these have parallels in 
the Kowy. NT Greek shows κερδῆσαι as well as κερδᾶναι, cvvaga as well 
as συναγαγεῖν, and ἡμάρτησα beside ἥμαρτον (cf. MGr jpapro=“ beg 
pardon!”); while προσέξαι can be illustrated from papyri.! Thumb 
goes on to say, “ While in general the ancient aorist has maintained its 
place, the present [MGr] stem has been quite frequently remodelled, and 
that on the basis of the aorist.” This process can be easily recognised in 
NT. The present suffix -yw, extremely productive in MGr, has made 
new present stems in δύνω (ἔδυσα), -xrév(v)o (-exrewa, like MGr σπέρνω: 
ἔσπειρα etc.), λιμπάνω (ἔλιπον), -xvv(v)o (-exvOnv), -στάνω (€orabnv—see 
§ 95): some of these began to appear in classical times. The simplifica- 
tion of present stems under an impulse from the aorist may be seen also 
in classical or Hellenistic exx. such as κυλίω (ἐκύλισα), virrw (ἔνιψα), 
ῥήσσω (ἔρηξα), στείλω (Ac 7°4—€oreida), σκέπτομαι (ἐσκεψάμην, which 
secured the victory of its present over the Attic rival σκοπέω), ἀμφιάζω or 
-ἐζω (ἠμφίεσα---866 § 33. 2), κρύβω (Lk 1°*?—see ὃ 9ὅ--ἐκρύβην), βλαστάω 
(ἐβλάστησα), ὀπτάνομαι (ὠφθην). 

2. Verbs in -aivw and -αίρω make weak aorist in -ava, -apa, without 
regard to the sound preceding this suffix : this is explained by Brugmann- 


1 Cf. παρέξασθαι CPR 175}5 (ii/a.p.) ad. 


§ 89] VERBS. 215 


Thumb Gr. 39 as due to the analogy of verbs in which the a was 
“pure.” Perhaps the working of this analogy, in the opposite direction 
to the general tendency of the Κοινή, may have been helped by the 
quality of the a vowel which kept its place in the rest of the verb. 

3. Xvv(v)o (the older yé) forms an abnormal aorist ἔχεα, best taken 
as a primitive strong aorist (¢ye- Fm, with weakened root in mid. ἐχύμην, 
ἐχύθης, ἔχυτο) : see Thumb in Brugmann Gram.* 676. This is the 
regular form in NT: on some ambiguities see the List, § 95. 


ACTIVE VOICE, 


Indicative— 
ἔλυσα ἐλύσαμεν ἔκρῖνα ἐκρίναμεν 
ἔλυσας ἐλύσατε ἔκρινας ἐκρίνατε 
ἔλυσε(ν) ἔλυσαν ἔκρινε(ν) ἔκριναν 


ΝΌΤΗ. — The infection of strong aor. endings is found in 2 sg. 
(from the influence of the common 3 sg. -e) in illiterate papyri of the 
Roman age, as P Oxy vii. 10675 (iii/A.D.) ἀφῆκες, 1. 119? (ii/ili A.D.) 
ἐποίησες 1} ἔπεμψες 18 ἔπλευσες, etc. : it is fixed in MGr. It appears in 
Rev 24 8*¥C ἀφῆκες, Mt 1135 D ἀπεκάλυψες, ἔδωκες Jn 177 NB ἔδωκες, E35, 
and even ἀφήκετε in Mt 237° B, Apparently it began in the Perfect, 
which accounts for its appearance at first mainly in -κα aorists: see § 92. 


Imperative— 
λῦσον λύσατε 
λυσάτω λυσάτωσαν 


Νοτε.---Ἰ. The MGr type δέσε, δέσετε is foreshadowed by some late 
forms in vernacular Koy: cf. Dieterich Unters. 248. Radermacher 
suggests that the middle λῦσαι, pronounced as λῦσε, may have started the 
assimilation to the present. 

2. There is no trace in Hellenistic of the 3rd pl. λυσάντων, which is 
regular in Attic until 300 B.c. (Meisterhans* 167). 


Subjunctive— 

λύσω λύσωμεν 

λύσῃς λύσητε 

λύσῃ λύσωσι 
Optative— 

λύσαιμι λύσαιμεν 

λύσαις λύσαιτε 

λύσαι λύσειαν OF -avev 


Norrt.—1. In papyri (Harsing 14) the 3 sg. is λύσειε(ν) more often 
than Avoa, but the exx. of -ee(v) are all from iii/a.p. or later (except 
one of ii/B.c.), and belong accordingly to the period in which the language 


216 A GRAMMAR OF NEW TESTAMENT GREEK. [δ 89-90 


of the schools gave the optative a short spell of renewed life. It seems 
that -ac was the suffix in the natural vernacular before this revival, and 
this is the only form evidenced in NT, unless we count Lk 6!! ποιήσειεν 
in NAW cu, Ac 1727 ψηλαφήσειεν in NE cu, which are ruled out by the 
demand fora 3 pl. Cf. Cronert Mem. 213, Meisterhans* 166 (“the so- 
called Aeolic forms in the aorist do not occur” in Attic imnserr.), 
Thackeray 215. 

2. In the pl. the evidence is somewhat conflicting. Lk 6!! BLA 1 ete. 
13 etc. 33 al. has ποιήσαιεν (ω -eav, NAW -eev), but Ac 1727 ABw 
ψηλαφήσ(ε)ιαν (see note 1), where only cu® have -aev. Here D reads 
-acav, Which Blass wanted to accept, mainly because it is regular in 
LXX (Thackeray 215). It must be remembered however that the LXX 
has other extensions of the 3 pl. -σαν which are not shared by NT, 
and may be dialectic variations : the MS attestation is not strong enough 
to force this form on a writer of Luke’s Greek culture. 


Infinitive— 
λῦσαι 


Note.—There is a strong tendency in vernacular Κοινή to sub- 
stitute the ending of the pres. inf. in act. and mid., so that the aor. infin. 
may be confused with the practically obsolete fut. infin. See Prol, 204 n.2, 
where it is noted that ἔσεσθαι is the only fut. inf. in NT except καταν- 
τήσειν Ac 267 B, εἰσελεύσεσθαι Heb 318, and χωρήσειν Jn 2125 NBC. 
Heb lc. is the only clear fut. here: the other two are probably mere 
aorists. The aor. inf. would carry the same meaning, and the -joa of 
the other MSS is best taken as a correction. 


Participle— 
λύσας -ασα -αν 
890. MIDDLE VOIOE. 

Indicative— 

ἐλυσάμην ἐλυσάμεθα 

ἐλύσω ἐλύσασθε 

ἐλύσατο ἐλύσαντο 
Imperative— 

λῦσαι λύσασθε 

λυσάσθω ες λυσάσθωσαν 
Subjunctive— 

λύσωμαι λυσώμεθα 

λύσῃ λύσησϑε 


λύσηται λύσωνται 


§ 90] VERBS. Oey, 


Optative— 
λυσαίμην λυσαίμεθα 
λύσαιο λύσαισθε 
λύσαιτο λύσαιντο 


ΝοΤΕ.---Εὐξαίμην is the only quotable instance, but the rest of the 
tense could have no other form. 


Infinitive— 
λύσασθαι 
Norr.—The substitution of -εσθαι, making the form identical with 


that of the obsolete future, is parallel with that noted under the active ; 
but there is no instance in NT. 


Participle— 
λυσάμενος 
PASSIVE VOICE. 

Indicative— 

ἐλύθην ἐλύθημεν 

ἐλύθης ἐλύθητε 

ἐλύθη ἐλύθησαν 
Imperative— 

λύθητι λύθητε 

λυθήτω λυθήτωσαν 


Notr.—The 2 sing. -r is for -6: by aspirate dissimilation. 


Subjunctive— 
λυθῶ λυθῶμεν 
λυθῇς λυθῆτε 
λυθῇ λυθῶσι(ν) 
Optative— 
λυθείην λυθείημεν 
λυθείης λυθείητε 
λυθείη λυθείησαν 


Norr.—A few instances occur in ΝΊ ---πληθυνθείη (1 Pet.1?, 2 Pet 15, 
Jude), λογισθείη (2 Ti 41°), τηρηθείη (1 Th 57°) — but the forms in 
Hellenistic are certain. Even the Atticisers hardly show the primitive 
(and Attic) short forms λυθεῖμεν -cire, though Moeris commended them. 
Cf. Scham Opt. bet Clem. Alex. 34; Harsing 22; whence it appears 
that the 3rd pl. -ciev survived where -eiyev -etre did not. 


218 A GRAMMAR OF NEW TESTAMENT GREEK. [8 90-91 


Infinitive— 
λυθῆναι 
Participle— 
λυθείς -etoa -έν 


D, FUTURE STEM. 


§ 91. For the formation of Future Stems see above, § 82. 


ACTIVE AND MIDDLE VOICES. 


From the earliest times in Greek the Future has a large 
proportion of Middle forms, there being whole categories of 
verbs in which a present active took a future middle without 
any ascertainable reason. On this subject, and on the assim- 
ilations which took place in Hellenistic, see Proleg. 154 f. 
Notes on the individual verbs will be found in the List. 

Verbs in -if{@ show some wavering between the -c- and 
the -e(c)- formation. Moeris (see Schweizer Perg. 178) 
makes -e?rae Attic and -ceras Hellenistic. The more normal 
form naturally secured a lead over its rival, which held its 
own perforce in the liquid verbs. All cases where the Con- 
tracted Future is found in NT will be noted in the List. 
WH (App? 170f.) make -ὦ 4, -cev habitually (exe. twice 
(δια)γκαθαριεῖ), -σομεν 323, -οῦσι except γνωρίσουσι, -σεται 
3295 5. -εἶσθε +. See the note above on the difference between 
LXX and NT in the future of verbs in -afw (§ 82). 


Indicative— 
Active. Middle. 

λύσω κρϊνῶ λύσομαι ἔσομαι κρινοῦμαι 
λύσεις κρινεῖς λύσῃ ἔσῃ κρινῇ 
λύσει κρινεῖ λύσεται ἔσται κρινεῖται 
λύσομεν κρινοῦμεν λυσόμεθα ἐσόμεθα κρινούμεθα 
λύσετε κρινεῖτε λύσεσθε ἔσεσθε κρινεῖσθε 
λύσουσι(ν) κρινοῦσι(ν) λύσονται ἔσονται κρινοῦνται 

Subjunctive— 


Alleged exx. of this imaginary mood are δώσῃ Jn 17%, 
Rev 88, ὄψησθε Lk 1378, which are only new aorists made 
from the future stem by the usual analogy. 


8. 91] VERBS. 219 


Optative— 

This mood, which in classical Greek only existed for one 
syntactical category, the representation of a fut. indic. in 
orat. obl. in past sequence, is entirely obsolete in Hellenistic, 
except for one or two artificialities of a late period. 


Infinitive— 
Active. Middle. 
λύσειν κρινεῖν λύσεσθαι ἔσεσθαι κρινεῖσθαι 


Nors.—This form can hardly be said to have any real vernacular 
existence; see ὃ 89. In Jn 2125 χωρήσειν was probably an aor., as far as 
the writer’s consciousness went ;! and the substitution of -ac@a in many 
places where so clear a future as ἐπελεύσεσθαι appeared in a formula 
shows that even this was felt as an aorist. Ἔσεσθαι is the one real 
exception, and even this only occurs in Ac: μέλλειν ἔσεσθαι (ter) is a set 
phrase, and 2380 μηνυθείσης δέ μοι ἐπιβουλῆς cis τὸν ἄνδρα ἔσεσθαι is in an 
official letter in stilted style. Outside Ac and Heb (and Jn 21?5—see 
above) the infin. is not found: οἵ. Heb 31%, Ac 267 B. 


Participle— 
Active. 


λύσων -ουσα -ov κρινῶν -otca -otv Avodpevos ἐσόμενος κρινούμενος 


Norr.—This also is very rare, but shows more signs of life than 
the infin. The only warrant for the contracted form is κατακρινῶν 
(Rom 8%), which might as well be present, and κομιούμενοι 2 Pet 21° 
ACw vg syr*! sah boh, which is certainly corrupt; but of course these 
forms would be used if the future pte. of a liquid verb were wanted. As 
before, the Lucan writings and Heb show the survival most. 


PASSIVE VOICE. 


The Strong Future (ἀλλαγήσομαι etc.) agrees exactly with 
the model of the Weak, which alone need be given. 


Indicative— 
λυθήσομαι λυθησόμεθα 
λυθήσῃ λυθήσεσθε 
λυθήσεται λυθήσονται 


ΝΟΤΕ.--Καυθήσωμαι in 1 Co 135 CK al. seems to be ἃ mere fusion 
of the vv.ll. καυθήσομαι and καυχήσωμαι----ἰΐ indeed we should take it as 
seriously as even this; it does not in any case provide us with a future 
subj. ! 


1 Blass’s remark about ‘‘ the spurious concluding verse” of Jn (G@r.? 202 n.) 
rests on no evidence at all: see Lake’s introduction (Cod. Sin. p. xx) for the 
supposed hostile witness of &*. 


220 A GRAMMAR OF NEW TESTAMENT GREEK. [8 91-92 


Infinitive— 
λυθήσεσθαι 
Norr.—Not in NT, but quotable 6... from P Tebt i. 61 (4)186 (land 
survey, B.C. 118). 
Participle— 
λυθησόμενος 
Notrr.—Found once (Heb 3°), and quotable from papyri. 


& 


Α΄. PERFECT STEM. 
ACTIVE VOICE. 


§ 92. The formation of the Strong Perfect Active is 
described above, ὃ 82 #; its frequent intransitive use, which 
caused it in old grammars to be counted as a middle, is noted 
in Prol. 154, with the theory that its unique person-endings 
betray a formation which in its prehistoric stages was neither 
active nor middle. 

Perfect stems with the o-gradation (historically identical with the 
proethnic Germanic vowel in sat, rang, wrote etc.) are still numerous in 
Hellenistic. Thus οἶδα, λέλοιπα, πέποιθα, γέγονα, ἐνήνοχα, πέπονθα. In 
the other vowel-series there are no traces left of the corresponding grada- 
tion, except εἴωθα compared with ἦθος. Thus εἴληφα has the same vowel 
as λήμψομαι, kéxpaya as κράζω, σέσηπα as σήπω. The roots with ev do 
not seem to preserve any o forms in the perfect: φεύγω makes πέφευγα, 
while rérvya keeps the weak gradation, originally characteristic of the 
plur.: cf. Eng. wrung, and the perfects begun etc., which were normal a 
few generations back. So γέγραφα, réraya, ἐλήλυθα. 

In two cases an old Perfect has produced in Hellenistic a new Present 
Stem : στήκω from ἕστηκα stand, and ypnyopéw from ἐγρήγορα am awake. 


The Person-endings in Hellenistic are levelled so as to 
be identical with those of the Weak Aorist in the indicative, 
except for the 3rd pl.: on this see below. The difference of 
gradation in the root of sing. and plur. no longer survives 
even in οἶδα. 

Strong and Weak Perfects may be taken together, their 
terminations being identical. 


Indicative— 
Perfect. 
οἷδα οἴδαμεν λέλυκα λελύκαμεν 
οἶδας οἴδατε λέλυκας λελύκατε 


οἶδε(ν) οἴδασι(ν) λέλυκε(ν) λελύκασι(ν) 


§ 92] VERBS. 221 


Norr.—1l. The old forms of οἶδα, 2 sg. οἶσθα and pl. ἴσμεν, ἴστε, 
ἴσασι(ν), were obsolete in vernacular Hellenistic. Moeris (ed. Pierson) 
p. 205 writes : “ἴσασι" ᾿Αττικῶς : οἶδασι᾽ κοινῶς, and cf. Phryn. (ed. Lobeck) 
p. 236 f., where instances of οἶδας from older Greek are cited.” It appears 
as early as B.c. 255 in P Petr ii. 4 (7)? (9)8, but οἶσθα in P Lille 114 
(mid iii/B.c.), an official letter, and even P Oxy viii. 1119 (a.D. 254) 
—cf. also Thackeray 278. Οἶδας appear 11 times in NT without variant, 
and the “regular” pl. very frequently in all persons, again agreeing with 
papyri: see Prol. 55, where Ionic is noted as the source of the levelled 
flexion. There is one certain ex. of the old forms, Ac 26* ἴσασι (no v.1.) : 
Heb 1217 ἴστε may also be a mark of this literary survival, appropriate 
in Heb, and in the speech of an educated man before a court. Cf. 
BGU i. 16314 (a.p. 108) also official. See further under Imper. 

2. Assimilation to the 8rd sg., on the lines of Impf. and Strong 
Aor., produced in the lower vernacular a 2nd sg. in -es, which passed 
on into the Weak Aor.: ὃ 89. It is rare in earlier papyri: see 
Thackeray 216, and some later exx. in CR xv. 36, xviii. 110, also Mayser’s 
note p. 321. In NT we find it plausibly read in Rev 25 AC κεκοπίακες, 
5 8% πέπτωκες, 1117 C cidnpes—also in Jn 857 B* ἑώρακες, Ac 217? B 
ἐλήλυθες, which last at any rate is highly improbable. 

3. In 3rd pl. the Weak Aor. -av secured a firm hold in the vernacular, 
being the last remaining difference between the aor. and perf. endings. 
(In some illiterate papyri the perf. -ao. invaded the aor.) It seems to 
begin in ii/B.c., and is found widely spread through the Κοινή : see Prol. 
52, Thumb Hellen. 170, Mayser 323f., Thackeray 212, CR xv. 36, 435, 
xviii. 110. In NT we find it in Ac 1680, Rom 167, Lk 950, Col 2! (see Prol. 
le.), with Rev 21° A, 195 NAP, 18% AC, Jn 177 ABCDL al., 17° BDLW, 
Jas 54 BP—too good a record perhaps to justify the suspicion I expressed : 
I must admit moreover that I aspersed unfairly the culture of some early 
papyri showing -av. 

4. The verb ἥκω “I have come,” which is a perfect in meaning, and 
by its κ suggests a formal connexion also with the perfect tense, developed 
a corresponding flexion in the pl. Thus ἥκαμεν P Par 489 (B.c. 158), 
ἥκατε P Grenf ii. 8618 (B.c. 95), ἥκασι Mk 88 SADW syrs latt al. (BLA 
boh substitute εἰσίν). Thackeray 269 and Mayser 372 show how well 
established this flexion is throughout, except in sing. indic. It is 
probably to be accepted in Mk Jc, : see Prol. 53, 


Pluperfect. 
dew ἤδειμεν (ἔγλελύκειν (ἐδλελύκειμεν 
ἤδεις ἤδειτε (ἐ)λελύκεις (ἐ)λελύκειτε 
ἤδει ἤδεισαν (ἐ)λελύκει (ἐ)λελύκεισαν 


Norr.—1. The past tense οἵ οἶδα has been assimilated to the other 
pluperfects. The sound of its initial vowel was in our period decidedly 
less removed from that of οἶδα than in Attic had been the case. 

2. The characteristic εἰ runs through the tense in NT forms. There 
are a few isolated irregularities in papyri: thus εἰώθησαν BGU i. 2507 


222 A GRAMMAR OF NEW TESTAMENT GREEK. _ [ὃ 92 


(ii/A.D.) (=Chrest. i. p. 114), εἰρήκης (pap. εἰρήκαις) P Par 3216 (B.c. 162), 
ὀμωμόκεμεν P Par 46! (8,6. 153). See Mayser 324. 
3. On the dropping of the augment in pluperf. see § 83. 


Imperative— 
ἴσθι ἴστε 
ἴστω ἴστωσαν 


ΝΟοΤΕ.--Ἰ. No perf. act. imper. forms occur in NT. In LXX we find 
such forms as κεκραγέτωσαν, πεποίθατε, πεποιθέτω (Job 125), 

2. It is best, except perhaps in Heb 1217 (see above, p. 221), to treat 
ἴστε as imper. wherever it occurs. In Jas 119 Mayor expresses a preference 
for indic., as also in Eph 5° and Hebl.c. But the only justification of 
this literary survival would be a clearly proved tendency in the author’s 
general style; and οἴδατε in Jas 44 matches the Greek of the writer. 
“Be sure of this,” Mayor’s alternative, is decidedly preferable ; and so 
in Eph l.c. (on which see Prol.® 245 also p. 22 f. above). 


Subjunctive— 
εἰδῶ εἰδῶμεν 
εἰδῇς εἰδῆτε 
εἰδῇ εἰδῶσι(ν) 


Nore. — The ordinary verb makes its subjunctive by combining 
participle and the verb εἶναι, as πεποιθὼς ὦ. 


Infinitive— 


εἰδέναι λελυκέναι ἑστάναι 


ΝΌΤΕ. --- The old strong perf. ἑστάναι occurs three times in NT, 
ἑστηκέναι Dever. 


Participle— 
εἰδώς -uia -ds λελυκώς ἑστώς -ὥσα -ds (gen. ἑστῶτος) 


ΝΟΤΕ.---Ἰ. On the gen. sing. fem. συνειδυίης see above, 88 37, 51. 

2. The strong ptc. ἑστώς, in simplex and compounds, maintains itself 
without serious challenge: it occurs in NT 57 times to 18 instances of 
ἑστηκώς, apart from places where the MSS are divided. Mk has 3:5, 
but all other writers use ἑστώς more frequently—except that Heb has 
ἑστηκώς in the one occurrence. The Lucan books show ἑστώς 22 times, 
and ἑστηκώς only twice; Rev has 9 ἑστώς against only 2. It is therefore 
no consequence of literary style one way or the other. 

3. Both ἑστηκώς (Rev 5° δὲ) and ἑστώς (Mk 1314 late uncials, Rev 14! 
046 and cu!!+) appear sporadically in neut. Since in both cases ad 
sensum construction would produce ἑστηκότα -@ra, this probably belongs 
to the general levelling of participial flexion : cf. § 65 (2). 


§ 93] VERBS. 223 


MIDDLE AND PASSIVE VOICE. 


§ 93. The flexion of this tense system depends on the 
character of the sound which ends the stem. Stems ending 
in a consonant have to use periphrastic 3 pl. in perf. and 
pluperf. indic.: the old forms with -atas -ato ( = -ntai -nto), 
like τετάχαται, were liable to be mistaken for 3 sg., and did 
not survive in the Κοινή. Hence types like πεπεισμένοι -αι 
-a εἰσίν, ἦσαν, had to supply the place. 

Vowel Guttural Labial Dental Liquid Nasal 
Stems. Stems. Stems. Stems. Stems. Stems. 
λελυ- τεταγ- γεγραφ- πεπειθ- ἐσταλ- μεμιαν- 


Indicative— 
Perfect. 
λέλυμαν τέταγμαι γέγραμμαι πέπεισμαι ἔσταλμαι μεμίαμμαι 
λέλυσαι τέταξαι γέγραψαι πέπεισαι ἔσταλσαι μεμίανσαι 


λέλυται τέτακται γέγραπται πέπεισται ἔσταλται μεμίανται 
λελύμεθα τετ ἄγμεθα γεγράμμεθα πεπείσμεθα ἐστάλμεθα μεμιάμμεθα 
λέλυσθε τέταχθε γέγραφθε πέπεισθε ἔσταλθε μεμίανθε 
λέλυνται Periphrastic 


Norgr.—1. The last column is not quite certain, and some persons do 
not occur. Tit 12° μεμιαμμένοις, Mk 3! 1150 ἐξηραμμένην, together with 
μεμαραμμένος and katnoxuppevos in Hermas and κατασεσημημμένα in P 
Oxy i. 1174 (cited by Blass) justify the pp against earlier on; but note 
συνκαθυφασμένα Is 328 αἱ. (Thackeray 224). The rest of the flexion may 
be assumed to be as in Attic. 

2. Stems in a, as rereAec-, are like the dental stems. 


Pluperfect. 
(ἔδγλελύμην (é)reraypnv (ἐ)γεγράμμην (ἐ)πεπείσμην ἐστάλμην 
(ὃλέλυσο (ἐγτέταξο (ἐγγέγραψο (ἐπέπεισο ἔσταλσο 
(ἐ)λέλυτο (ὃτέτακτο (ἐγγέγραπτο (ἐ)πέπειστο ἔσταλτο 


(δλελύμεθα (ἔγτετάγμεθα (ἐ)γεγράμμεθα (ἐ)πεπείσμεθα ἐστάλμεθα 
(ἐ)λέλυσθε (ὃτέταχθε (ἐγγέγραφθε (ἐ)πέπεισθε ἔσταλθε 
(ἐ)λέλυντο Periphrastic 

Norr. — Some of these forms are inferred. No form of the 
(€)pepudupny type occurs. 


Imperative— 
λέλυσο λελύσθω λέλυσθε λελύσθωσαν 


Nots.—The form πεφίμωσο, from φιμόω muzzle, can be quoted. 
The tense is very rare: if speakers of the Κοινή had occasion to use it 


224 A GRAMMAR OF NEW TESTAMENT GREEK. [§ 93-95 


they presumably used the old forms, which can be inferred from the 
flexions given. 


Subjunctive— 
Like the Optative, this was periphrastic (λελυμένος ὦ 
etc.) in earlier and later Greek. 


Infinitive— 
λελύσθαι τετάχθαι γεγράφθαι πεπεῖσθαι ἐστάλθαι 


Participle— 
λελυμένος τεταγμένος γεγραμμένος πεπεισμένος ἐσταλμένος μεμιαμμένος 


μι VERBALS. 


§ 94. One form of the gerundive in -τέος can be quoted 
from NT, viz. βλητέον from βάλλω (Lk 5**): see Prol. 222. 
In form it agrees with the verbal in -τός, the meaning of 
which is discussed in Prol. 221 f. 

Papyrus instances of the gerundive are P Giss 1. 40 +? (a.p. 215) διὰ 
τοῦτο οὔκ εἰσιν κωλυτέοι, P Tebt i. 61 (Ὁ) 2395. (8.6. 118—a land survey) 
εἰ [α]ὐτὴ [ἀνταναι]ρετέα [ἄλλη δὲ] ἀπὸ ὑπολόγου ἀνταναιρεθεῖσα ἀποκα- 
ταστατέα (the whole formula elsewhere), P Par 63 (ii/B.c.) °° χρηστέον, 
119 ἐνγραπτέον; Wels μεριστέον. These are all official, and in themselves 
inadequate warrant for really vernacular use. 


Verbal adjectives in -τός have recessive accent when 
compounded. There are about 150 of them in NT, and the 
formation seems to be still living, so that it can be made 
from new verbs. The -τός suffix in non-derivative verbs was 
originally added to the weakened root, as we may see in 
θετός, πιστός, φθαρτός, -υτὸς (ἀπρόσιτος): contract verbs add 
it to the long vowel seen in future, as ἀγαπη-τός, ἀλάλη-τος, 
μισθω-τός, and other derivative verbs to the bare stem, as 
βδελυκ-τός, συτισ-τός, ete. 


LIST OF VERBS. 


§ 95. This list includes all verbs occurring in NT (or in good 
MSS thereof), except— 
(1) regular Contracta. 
(2) regular verbs in -yw and -afw, and those in -ἰζω which show 
no future act. or mid. 


ὃ 95] VERBS. 225 


(3) verbs with no forms outside the present stem, and with 
nothing noteworthy to record. 

“ Regular” reduplication of verbs with initial o implies σεσ- 
where a vowel follows, ἐσ- where a consonant. 

Verbs are set down in the simplex, preceded by a hyphen when 
the form only occurs in a compound. The list of quotable 
compounds is appended in each case. A few Compound 
Verbs are given in their alphabetic place for points affect- 
ing the preposition. The prepositions are given in the 
assimilated form for the present indicative, according to 
classical orthography. 

Roman numerals attached to the present stem denote conjugation 
classes. 

An obelus denotes a form apparently not older than the Hellen- 
istic age. When placed on the extreme left it means that 
the whole verb is late. Suppletives are enclosed in square 
brackets. 

The indicative form stands for anything occurring in the tense 
paradigm. Occasionally the mood form is quoted for 
special reasons. 

The regular tense-formations of verbs not included in this List 

(see above) are as follows :— 


PRESENT. Weak Aorist. FUTURE. ° PERFECT. 
(1) -ἄω (a pure) -aca τᾶσω -ᾶκα 
-ἀσάμην τἄάσομαι -ἅμαι 
τἄθην -αθήσομαι 
-aw (a impure)\ τησα ete. τήσω ete, -nka ete. 
; ι 
“EW 
τόω -woa ete. -ώσω ete. τωκα ete. 
-ἰζω -ισα etc. [in List | -UKa. 
τίσθην τισμαι 
(2) -vw (αύω etc.) τ-οσα ete. -vow etc. -vka etc. 
-alw -ἄσα -ἄσω -aKa 
τἄσαμην -ἄσομαι πασμαι 
-άσθην -ασθήσομαι 


A. Present. 8. STRONG C. Weak OD. Furure. #. PERFEcr. 


AORIST. Aorist. 
T ἀγαλλιάω (Υ 11.) ἠγαλλίασα 
exult ἠγαλλιασάμην 
ἀγαλλιῶμαι ἠγαλλιάθην : 


ἠγαλλιάσθην (Jn ὅ55 BL) 


226 A GRAMMAR OF NEW TESTAMENT GREEK.  [§ 95 


‘A Hellenistic variation on ἀγάλλεσθαι after ναυτιᾶν, κοπιᾶν 
ἀγωνιᾶν," etc. (Blass-Debrunner). 


ἀγγέλλω (VIL) -ἠγγέλην T ἤγγειλα -ἀγγελῶ -ἤγγελμαι 
announce ἠγγειλάμην 
-ἤγγελλον 
Comp. ἀν-, ἀπ-, δι-. ἐξ-, ἐπ-, κατ-, παρ-, προ-επ-, προ-κατ- 
-ἀγνυμι(11. β.6ὴ)-ἐάγην -ἔαξα -ἐάξω ἡ 
break Subj. κατεαγῶσιν fF See ὃ 83 (1). 
Jn 1951: see § 83 (1). 
Comp. κατ- : 
-ἄγχω (I. a) ἠγξάμην 
choke 
Comp. ἀπ- 
ἄγω (I. 9) ἤγαγον -ἦξα ἄξω ἦγμαι 
lead Not Attic, but ἀχθήσομαι 
ἦγον old. Found 
ἄγομαι in illiterate 
ἠγόμην papyri. 
VERBAL -ἀκτός ἤχθην 


Comp. ἀν-, ἐπ-αν-, ἀπ-, συν-απ-, δι-) εἰσ-, παρ-εισ-, ἐξ-, ἐπ-, κατ-, 


μετ-, παρ-; περι-, προ-, προσ-, συν-, ἐπι-συν-, ὑπ- 


αἰνέω (Υ 11.) πἥνεσα καἰνέσω 
praise 
᾿Επαινέσω in 1 Co 


1122.53 } prob. 


-vouv 
VERBAL -aivetos 


Comp. ἐπ-, Tap- aor. subj. 
aipéw (VII.) -εἷλον (ἑλεῖν) “peOnv ENO Ὁ τῆρημαι 
take εἱλόμην (LXX and On the general 
papp.— Κοινή spell- 
-αἱροῦμαι from ing -εἴρημαι, 
-ἡρούμην (-α forms : ὃ 880). aor.). see above, 
αἱρήσομαι § 83 (6). 
-αἱρεθήσομαι 
VERBAL -αἱρετός 
Comp. ἀν-, ἀφ-, δι-, ἐξ, καθ-, περι-, προ- 
αἴρω (VII) ἦρα (ἄραι) ἀρῶ ἦρκα 
raise ἤρθην ἀρθήσομαι Appar 


αἴρομαι 
Comp. ἀπ-, ἐξ-, ἐπ-, μετ-, συν-, ὗπερ- 


§ 95] VERBS. 227 


From Fapyw: not contracted 
from cognate deipw. The 
aor. must not be written 
with « subscript.: 


αἰσθάνομαι ἤσθόμην' 
(11. α) 


perceive 


-aicxtvw (VIT.) ἠσχύνθην αἰσχυνθήσομαι 
shame (-αισχ. : see ὃ 83 (6)) 
αἰσχύνομαι 
ἠσχυνόμην 
VERBAL -αἰσχυντός 
Comp. ἐπ--, κατ-- 


ἀκούω (Υ 11.) ἤκουσα ἀκούσω Γἠ ἀκήκοα 
hear ἠκούσθην ἀκούσομαι 
ἤκουον ἀκουσθήσομαι 
ἀκούομαι 
Comp. δι-, εἰσ-, ἐπ-, παρ-, προ-, ὑπ- 
ἀλείφω (I. a) ἤλειψα «ἀλείψω 
anoint ἠλειψάμην 
ἤλειφον «ἠλείφθην 
Comp. ἐξ- (Ac 819 ἐξαλι- 
φθῆναι WH: 
ἁλίσκομαι : see ἀναλίσκω see § 35). 
VERBAL -ἁλωτός 
-ἀλλάσσω (VII.) -ἠλλάγην ἥἤλλαξα ἀλλάξω -ἤλλαγμαι 
change ἀλλαγήσομαι 
-ἤλλασσον 
-ἀλλάσσομαι 
ComP. ἀπ-, Sts, κατα; ἀπο-κατ-, μετ-, συν- 
ἅλλομαι -ἡλόμην 
leap (-a forms: see 
§ 88). 
Comp. ἀν-, ἐξ-, ἐφ- 
ἁμαρτάνω( 1. α) ἥμαρτον ἡμάρτησαΪ ἠ ἁμαρτήσω ἡ ἡμάρτηκα 
sin (see § 89). 
VERBAL -ἁμαρτητός 
Comp. προ- 
ἀμύνομαι (VIT.) ἠμυνάμην 


take revenge 
VOL, Il. PART I1,—16 


228 A GRAMMAR OF NEW TESTAMENT GREEK. [8 95 


ἀμφιέννυμι (11. β. 6) ἠμφίεσμαι 
ἀμφιάζω t (VIL) (See $83 (7)) 
clothe 


The simplex ἕννυμι had long been obsolete : 
so was the II. 8 present-stem except in 
semi-literary language. On the variant 
Κοινή presents -άζω and -έζω, both found 
in good uncials of Lk 1278, see Voca- 
bulary, s.v., and supra, p. 68. 


ἀναλίσκω (IV.) τς ἀνήλωσα ἀναλώσω 
ἀναλόω ἡ (Υ11.) (ἀναλῶσαι) 
spend, destroy ἀνηλώθην 


Comp. κατ-, προσ- 

A very early compound (ἀνα-ξαλίσκω), with afa contracted to ἃ: 
cf. ἁλίσκομαι. The late “regular” present (2 Th 28 x*) is a 
back-formation from the future and aorist. 


ἀνοίγω (I. a) Avolyny tT ἀνέῳξα ἀνοίξω ἀνέῳγα 
open ἤνοιξα ἀνοιχθήσομαι F ἀνέῳγμαι 
ἀνοίγομαι ἠνέῳξα ἡ ἀνοιγήσομαι T ἠνέωγμαι T 
-ἤνοιγον (ἀνοῖξαι) ἤνοιγμαι T 
Comp. δι- ἀνεώχθην 
See Vocabulary, s.v. ἠνοίχθην T 


For the strange irregulari- ἠνεῴχθην ἡ 
ties of augment and re- (ἀνεῳχθῆναι) 
duplication, see § 83 (1). 

The simplex otyw or 
οἴγνυμι (on which see 
Brugmann* 8510 n.) 
was extinct in Hellen- 


istic. 
-ἀντάω (VIT.) -ἤντησα -ἀντήσω F -ἤντηκα 
meet (cl. -ομαι) 


Compe. ἀπ-, kat-, συν-, ὑπ- 
The simplex became obsolete early. 


ἅπτω (VIT.) ἥψα 
grasp, kindle ἡψάμην 
ἅπτομαι ἤφθην 
ἡπτόμην 


Compe. ἀν-, καθ-, περι- 


§ 96] VERBS. 229 


ἀρέσκω (IV.) 
please ἤρεσα ἀρέσω 
ἤρεσκον 
VERBAL ἀρεστός 


ἀρκέω (Υ11.) ἤρκεσα ἀρκέσω 
suffice ἀρκεσθήσομαι T 
VERBAL ἀρκετός 
Comp. ἐπ- 
ἁρμόζωϊ (VIL) ἡρμοσάμην 
Jit 
The Attic pres. was ἁρμόττω. 
ἁρπάζω (VII.) ἡρπάγην T ἥρπασα ἁρπάσω + -«ἡρπάκειν 
seize ἡρπάσθην ἁρπαγήσομαι f 
see Rutherford 
NP 407. 


Comp. 8t-, συν- 
On the mixture of guttural and dental stem see Proleg. BBs also 


Brugmann Gir.* 359. 


ἄρχω (I. a) ἠρξάμην ἄρξομαι 
be first 
ἄρχομαι 
τἦρχον 
ComP. év-, προ-εν-, ὗπ-, προ-υπ- 
ἀστράπτω ( 11.) -ἥἤστραψα 
lighten 
Comp. ἐξ-, περι- 


αὐξάνω (II. a) ηὔξησα αὐξήσω 
αὔξω ([11.) ηὐξήθην 
WAL : 
ηὔξανον 
αὐξάνομαι 
Comp. συν-, ὑπερ-(αυξάνω) 


ἀφίημι, imperf. ἤφιον : for other forms see under -ἴημι. 


-Baivw (VII.) -έβην (§ 88) -βήσομαι -βέβηκα 
go 
-ἔβαινον 
CausaL -βιβάζω (ἀνα-, ἐμ-, ἐπι-, κατα-, mpo-, συν-), conjugated 
regularly. 


VERBAL -βατός 


230 A GRAMMAR OF NEW TESTAMENT GREEK. [890 


Comp. ἀνα-, ἀπο-, δια-, ék-, ἐμ-ν, ἐπι-, κατα-, μετα-, παρα-, προ-, 
προσ-ανα-, συμ-, συγ-κατα-, συνπανα-, ὕπερ- 

The simplex was very nearly extinct in common speech when the 
Κοινή arose. See Vocabulary, 8.0. 


βάλλω (VIT.) ἔβαλον ἐβλήθην βαλῶ βέβληκα 
cast -ἐβαλόμην -βαλοῦμαι ᾿-βεβλήκειν 

βάλλομαι (For -a forms βληθήσομαι βέβλημαι 
-ἔβαλλον see § 88.) ἐβεβλήμην 


VERBALS -βλητός, βλητέος 
Comp. ἀμφι- ἀνα-, ἀντι-, ἀπο-, δια-, ἐκ-, ἐμ-, ἐπι-, κατα-, μετα-, 
παρα-, παρ-εμ-, περι-, προ-, συν-, ὕπερ-, ὗπο- 


βαπτίζω (Υ11.} ἐβάπτισα βαπτίσω βεβάπτισμαι 
baptize ἐβαπτισάμην βαπτισθήσο- 
βαπτίζομαι ἐβαπτίσθην μαι 
ἐβάπτιζξον 
ἐβαπτιζόμην 
βάπτω (VIL) ἔβαψα βάψω βέβαμμαι 
dip (Rev 1918 A) 
Comp. ἐμ- 
Bapéw ἵ (VIT.) -ἐβάρησα ἵ βεβάρημαι 
burden ἐβαρήθην ἡ 
βαροῦμαι 


(ὋΜΡ. ἐπι-, κατα- 

The verb was apparently a back-formation from the perf. pass, 

The alternative βαρύνω is found in Lk 21% βαρυνθῶσι DH and 
Mk 14? καταβαρυνόμενοι. 


βασκαίνω (VIT.) ἐβάσκανα 
bewitch 
βαστάζω (VII.) ἐβάστασα βαστάσω 
carry 
ἐβάσταζον 


VERBAL -βαστακτός 
The guttural forms are not found in NT exe. in verbal and in 
Rev 23 βαστάξαι P 1, 38: they are frequent in papyri. 


βδελύσσομαι (VIT.) ἐβδέλυγμαι 
loathe 
VERBAL βδελυκτός 
BiBpdcxw(IV.) BéBpwxa 
eat 


VERBAL -βρωτός 


§ 96] . VERBS: ΘΠ 


βιόω (11.) ἐβίωσα 
live supplanted strong 
Supplied in earlier Greek δου. ἐβίων. 
an aorist to ζήω. 


βλάπτω (VIT.) ἔβλαψα 
hurt 

βλαστάνω (II. a) ἐβλάστησα 

βλαστάω t(VIT.) Perhaps Ionic (Hippocrates) : re- 
grow places strong aor. €BAacrov. 


Βλαστᾷ (Mk 427) is an instance of confusion between the two 
classes of Contracta which make fut. in -jow (see ὃ 84): the 
back-formation BAacréw is perhaps quotable even in Aeschylus. 


βλέπω (I. a) ἔβλεψα βλέψω ἡ 
ἔβλεπον, look -ἐβλεψάμην { Herodotus has ἀναβλέψω. 
βλέπομαι 


Compe. ἀνα-, ἀπο-, δια-, ἐμ-, ἐπι-, περι-, προ- 

The simplex appears once in aor. and once in fut., over hundred 
times in present stem, as the suppletive of εἶδον (so βλέπω, εἶδα 
in MGr). 

βούλομαι (II. a) ἐβουλήθην 
wish 
ἐβουλόμην 

Blass’s statement (cf. Gramm.! § 66%, also p. 58) that this verb is 
“taken from the literary language,” fits badly with its abun- 
dance in the papyri: see Vocabulary, s.v. On its augment see 
§ 83 (1), on βούλει, ὃ 85. 


βρέχω (I. a) ἔβρεξα 
wet, rain 
γαμέω (VIT.) ἔγημα γεγάμηκα 
marry ἐγάμησα Tt 
ἐγάμουν ἐγαμήθην 


The use οὗ γαμοῦμαι = nubo is obsolete, except occasionally in legal 
documents: see Prol. 159. Cf. the derivative present stems 
γαμίζω t and γαμίσκομαι. 

γελάω (VIT.) yedaow T 
laugh 
-ἐγέλων 
Comp. κατα- 
γηράσκω (IV.) ἐγήρασα 
grow old Trans. in older Greek, as against 
strong aor. 


232 A GRAMMAR OF NEW TESTAMENT GREEK. [ὃ 95 


γίνομαι (I.a.a) ἐγενόμην ἐγενήθην γενήσομαι γέγονα 
become For -α forms (ἐ)γεγόνειν 
ἐγινόμην see § 88. γεγένημαι 


Comp. ἀπο-, δια-, ἐπι-, παρα-, συμ-παρα-, προ- 

The older form γίτγντ-ομαι passed phonetically into yw. in Τοηϊο--- 
see Thumb, Gr. Dial, p. 352—and thence spread in the Κοινή. 
It is most frequently γείνομαι in MSS: see WH App.? 160. 
W still shows γιγν. sometimes. 


γινώσκω ([V.) ἔγνων (see ὃ 88)ἐγνώσθην γνώσομαι ἔγνωκα 
perceive γνωσθήσομαι ἐγνώκειν 
γινώσκομαι ἔγνωσμαι 
ἐγίνωσκον 


VERBAL γνωστός 

Comp. ἀνα-, δια-, ἐπι-, κατα-, προ- 

For γι-γνώ-σκω (Attic), as with γίνομαι above, and like it spelt 
yew. (Τγν. in W sometimes). 


γνωρίζω ( VIL.) ἐγνώρισα γνωριῶ (Col 
make known ἐγνωρίσθην 4° "ἢ 
γνωρίζομαι γνωρίσω ἡ 
Comp. éva-, δι- (only Lk 217 APo) 
Tyoyyvtw(VII.) ἐγόγγυσα 
grumble 
-ἐγόγγυΐον 
Comp. δια- 
γράφω ἐγράφην ἔγραψα γράψω γέγραφα 
write -ἐγραψάμην γέγραμμαι 
γράφομαι -ἐγεγράμμην 


VERBAL γραπτός Ξ 
ComP. ἀπο-, éy-, ἐπι-, κατα-, προ- 
δεῖ (1. α) 
impers. must 
ἔδει 
The only other forms occurring are δέῃ subj., and δέον δέοντα pte. 
See δέομαι. 


Setkvupe (II. β. ὃ) ἔδειξα δείξω δέδειγμαι 
show -ἐδειξάμην 
δείκνυμαι ἐδείχθην 


Comp. ἀνα-, ἀπο-, ἐν-, ἔπι-, ὕπο- 
For forms as from δεικνύω, see § 86. 
δέομαι ἐδεήθην 
need 
ἐδεόμην 


§ 95] VERBS. 233 


Comp. προσ- 
For flexion see § 85. The active forms the impersonal det. 


Sépw (I. a) ἔδειρα δαρήσομαι 
beat 
δέχομαι (I. a) ἐδεξάμην δέξομαι δέδεγμαι 
receive -ἐδέχθην 
-ἐδεχόμην 


VERBAL δεκτός 
Comp. ἄνα-, ἀπο-, ἀπ-εκ-, δια-, εἰσ-, ἐκ-, ἐν-, ἐπι-) παρα-, προσ-, ὕπο- 


δέω (11.) ἔδησα δήσω δέδεκα 
bind -ἐδησάμην δέδεμαι 
ἐδέθην -ἐδεδέμην 


Comp. κατα-; περι-, συν-, ὕπο- 


The present stem (inflected like φιλέω) is not found in NT. 


διακονέω (VIT.) διηκόνησα 7 διακονήσω 
minister (inf.) διακονη- 
διακονοῦμαι θῆναι 
διηκόνουν 
Attic used doubly augmented forms, ἐδιηκόνουν, etc. 
διδάσκω (1V.) ἐδίδαξα διδάξω 
teach ἐδιδάχθην 
ἐδίδασκον 


VERBAL διδακτός 


δίδωμι (1. α. ὃ) (pl.) ἔδομεν ἔδωκα δώσω δέδωκα 
give -ἐδόμην (subj.) δώσω 1 -δώσομαι δέδομαι 
δίδομαι (§ 91) δοθήσομαι (ἐ)δεδώκειν 
ἐδίδουν ἐδόθην 
-ἐδιδόμην 


VERBAL -δοτός 

Comp. ἀνα-, ἀντ-ἀπο-, ἀπο-, δια-, ἐκ-, ἐπι-, μετα-, παρα-, προ- 

See for flexion, and for later thematic forms in present stem, 

§§ 86-88. 
διψάω (VII.) ἐδίψησα διψήσω 
thirst 
Inserted here because no longer from stem διψη- (inf. διψῆν) as in 

el. Gr. It is now like τιμάω. 


διώκω (I. a) ἐδίωξα διώξω δεδίωγμαι 
pursue διωχθήσομαι 
διώκομαι 


ἐδίωκον 


234 A GRAMMAR OF NEW TESTAMENT GREEK. [8 95 


Comp. ék-, κατα- 
“δρεμ, run—see under τρέχω, to which it acts as suppletive. 


δοκέω (VIT.) ἔδοξα 
suppose 
ἐδόκουν 
δύναμαι (IT. a. δ) ἠδυνήθην δυνήσομαι 
can ἠδυνάσθην 
ἐδυνάμην The latter was Ionic: the 
ἠδυνάμην : two forms represent vary- 
On the augm. see ὃ 83 (1). ing dialects contributing 
VERBAL δυνατός to the Κοινή. 
δύνω (11.) ἔδυν (§ 88) ἔδυσα -δέδυμαι 
-δύω -ἐδυσάμην 
plunge 
-δύομαι 


Comp. ἀπεεκ-, ἐκ-, ἐν-, ἐπι-, ἐπ-εν-, παρ-εισ- 
From the same root the new present stem ἐνδιδύσκω (1Υ̓.). 7 
édw (VII.) εἴασα ἐάσω 
allow ᾿ 
εἴων 
Comp. προσ- 
ἡ ἐγγίζω (VIL) ἤγγισα ἐγγιῶ ἤγγικα 
approach (éyyiow Jas 48 A) 
ἤγγιζον 
Comp. προσ- 
A Κοινή verb, starting in Polybius. 


ἐγείρω (VIT.) ἤγειρα ἐγερῶ ἐγήγερμαι 
rouse ἠγέρθην ἐγερθήσομαι 
ἐγείρομαι 


-ἐγειρόμην (§ 88 (4)) 
Comp. δι-, ἐξ-, ἐπ-, συν- 


ἡ ἐδαφίζω (VIL.) ἐδαφιῶ 
raze 

ἐθίζω (VII.) εἴθισμαι 
accustom 

ἔθω εἴωθα 
be wont εἰώθειν 

The present stem only occurs in pte. (twice in Homer). 

* Fed εἶδον (-a, see ὃ 88) εἰδήσω οἶδα (§ 92) 

perceive ἤδειν 


Comp. ἀπ-, ἐπ-, προ-, συν-, ὕπερ- 


§ 95] VERBS. 235 


This root forms no present stem: εἶδον is used as 
aor. to βλέπω, and οἶδα makes a separate verb. 
* ξεικ ἔοικα 
resemble 
Like ,/reé, this root forms no present stem. The present only 
occurs in one NT writer (Jas 1 33), 
-εἴκω (I. @) εἶξα 
yreld 
Comp. ὑπ- 
-εἶμι (1. δ) 
go 
-ἥειν 
VERBAL -ιτός 
Comp. ἄπ-, εἴσ-, ἐξ-, ἔπ-, σύν- 
For the flexion of the few surviving forms, see ὃ 86. 


εἰμί (1. δ) ἔσομαι 
be 3 sg. ἔσται, other- 
ἤμην t wise regular. 


Comp. ἄπ-, év=, ἔξ(εστι), πάρ-, συμ-πάρ, σύν- 
For the flexion see § 86. 


eipw (VIT.) ἐρρέθην (inf. ἐρῶ εἴρηκα 
say ῥηθῆναι, pte. εἰρήκειν 
VERBAL ῥητός ῥηθείς). εἴρημαι 
Comp. προ- The ε (Ionic) is only found 
in indic. 


The present stem in use is λέγω, g.v.; the aor. act. εἶπον -- 568 
*Fex-. The present εἴρω is obsolete early. 
ἐλαύνω (VIT.) -ἤλασα ἐλήλακα 
drive 
ἐλαύνομαι 
ἠλαυνόμην 
Comp. ἀπ- 
The root is éAa-: the difficult present stem is held by Brugmann 
(Gramm. 221) to be a denominative from a noun *éAavvos. 


ἐλέγχω (I. a) ἤλεγξα ἐλέγξω 
convict ἠλέγχθην 
ἐλέγχομαι 
Comp. δια-κατ-, ἐξ- (Jude!, a few cursives only). 
ἐλεέω \ (VIL) ἠλέησα ἐλεήσω ἠλέημαι 
ἐλεάω T ἠλεήθην ἐλεηθήσομαι 
pity 


On the variation in present stem see § 84. 


236 A GRAMMAR OF NEW TESTAMENT GREEK. [ὃ 95 


ἑλίσσω (VIT.) ἑλίξω 
roll up 
ἐλίσσομαι 
The (“ poetic, Ionic and late prose”) pres. εἰλίσσω (= ἐξελίσσω) 
is found in Rev 61 P cu™: Blass (Kiihner ii. 417) shows that 
spir. lenis is older, but the analogy of ἕλιξ, etc., produced ἑ. in 
later times. 


ἑλκόω (VIT.) εἵλκωμαι 
make sore 
ἕλκω (1. a) εἵλκυσα ἑλκύσω Τ 
drag The addition -voa is due to the analogy 
εἷλκον of the synonym eipvoa: in its turn 
Comp. é€- it produced a new future. 
t ἐλλογάω (Υ 11.) 
impute 


ἐλλογοῦμαι and -ὥμαι 
On the variation in present stem see ὃ 85. 


ἐλπίζω (VIT.) ἤλπισα ἐλπιῶ F ἤλπικα 
hope 
ἠλπιΐον 
ἐλπίζομαι 
Comp. ἀπ- (ἀφ-, see p. 98), προ- : 
Veitch notes that ‘‘the early Greeks . . . were chary in express- 
ing confidence in the future.” ᾿Ελπιῶ is a late form built on 
Attic models, 


ἐμβριμάομαι ἐνεβριμησάμην 
and -έομαι (VIT.) ἐνεβριμήθην 7 
groan 
Cited for varying present stem: see § 85, pp. 198, 201. 
ἐμέω (1. a) ἤμεσα 
vomit 
Probably a very early thematising of Ἐξέμεμι (Skt. vdémime). 
ἐμφανίζω (VIT.) ἐνεφάνισα ἐμφανίσω 
manifest ἐνεφανίσθην 
A denominative from ἐμφανής : on the place of the augment see 
§ 83 (7). 
ἐπιορκέω ( VII.) ἐπιορκήσω 
perjure edu. Mt ὅ88 &: see p. 99. 


ἐπίσταμαι (1. 8) 
understand 


§ 95] VERBS. 237 


For flexion see ὃ 87. In NT only in present stem. It is an 
aorist-present from the middle of στῆναι. 


*rer speak εἶπον (-a: ὃ 88). 
Comp. dvt-, ἀπ-, προ- 
The aor. (a reduplication, the εἰ constant accordingly in the 
moods) is used for λέγω and the family of (εἴρω), q.v. 


-ἔπομαι (1. a) 
Sollow 
-εἱπόμην 
Comp. συν- 


ἡ ἐραυνάω (VIT.) ἠραύνησα 
search ; 
VERBAL -ἐραυνητός 
Comp. é€- 
The older form ἐρευνάω occurs in the mass of MSS: see § 37 
and Prol. 46 n.?. 


ἐργάζομαι ( VII.) ἠργασάμην (eipy.) εἴργασμαι 
work -εἰργάσθην 
ἠργαζόμην 


Comp. κατ-, περι-, προσ- 


On the augment see § 83 (3). 


ἐρείδω (1. a) ἤρεισα 
steck fast 


ἐρεύγομαι (1. a) ἐρεύξομαι 
utter 


ἐρίζω (VII.) ἐρίσωΤ 
strive 


ἑρμηνεύω (VII.) -ἑρμήνευσα 
interpret On dropped aug- 
ἑρμηνεύομαι ment see ὃ 83 (4). 
VERBAL -ἑρμηνευτός 
Comp. δι-, μεθ- 


ἔρχομαι (1. α) ἦλθον (-α, § 88) ἐλεύσομαι ἐλήλυθα 
come ἐληλύθειν 
ἠρχόμην ἵ (διέρχοντο Jn 208 DA*—see ὃ 83 (4)) (δ 83 (4)) 


VERBAL -ἦλυτός 

Comp. dv-, dvti-map-, ἀπ-, δι-, δι-εξ-, εἰσ-, ἐξ-, én, ἐπτ-αν-, ἐπ-εισ-, 
KQT=, παρ-, παρ-εισ-, περι-, προ-, προσ-, συν-, συν-εισ- 

The conjugation is made up from three roots, ἐρθ, ἐλυ and ἐνθ 


238 A GRAMMAR OF NEW TESTAMENT GREEK. [§ 95 


(Doric ἦνθον), which have influenced one another’s forms. On 
Attic usage of present and future stems, see Rutherford VP 


108 ff. 
ἐρωτάω (-ewt) (VII.) ἠρώτησα ἐρωτήθω 
ask ἠρωτήθην 
ἠρώτων (-ουν +) On the present stem see ὃ 84, 
Comp. 8t-, ἐπ- 
ἐσθίω (VII.) [ἔφαγον] [φάγομαι 17 
ἔσθω (VI.) Κοινή form, made (like 
eat φύγομαι and λάβομαι) 
ἤσθιον on the analogy ἔπιον: 
Comp. κατ-, συν- πίομαι ::ἔφαγον : φάγο- 


μαι ---- 8ι0 Brugmann 
Gr.4 383. Cf. § 85 
above. 
"Eo Ow (whence ἐσθίω by addition of a further suffix) is as old as 
Homer: it appears five times in ptc. and once (Lk 2230) in 
subj. The suppletive τρώγω is used in present stem. 


ev- For augment of verbs in εὖ- see 8 83 (6)—the variants will not 
be noted here, forms being given as in WH. 


evayyedifw tf (VIT.) εὐηγγέλισα ἡ εὐηγγέλισμαι 
evangelise εὐηγγελισάμην 
εὐαγγελίζομαι εὐηγγελίσθην 
εὐηγγελιζόμην 
Comp. προ- 
j edapeotéw (VIT.) εὐαρεστῆσαι εὐαρέστηκα 
please (einp. — see 
εὐαρεστοῦμαι § 83 (7)). 
εὐθύνω (VIT.) εὔθυνα (in moods) 
straighten 
Comp. κατ- 
εὐοδόω (VII.) εὐοδωθῶ εὐοδωθήσομαι 
give a fair way (1 Co 16? AC ai.) 
εὐοδοῦμαι 
On Hort’s proposal to read the pres. subj. εὐοδῶται, as perf. see 
§ 83 (5). , 
εὑρίσκω (IV.) εὗρον (-a εὑρέθην εὑρήσω εὕρηκα 
πα forms see § 88). εὕὑρεθήσομαι 


εὑρίσκομαι εὑρόμην 
ηὕρισκον (εὗρ.) 
ηὑρισκόμην 

Comp. ἀν- 


§ 95] VERBS. 239 


εὐφραίνω (VIL) ηὐφράνθην εὐφρανθήσομαι 
gladden 
εὐφραίνομαι 
εὐφραινόμην 
εὔχομαι (1. α) εὐξάμην -εὔξομαι 
pray 
ηὐχόμην (εὐ.) 
Comp. προσ- 


ἔχω (I. α) ἔσχον ἕξω ἔσχηκα 
have ἐσχόμην ἕξομαι 
ἔχομαι 
εἶχον 
εἰχόμην 


VERBAL -ἑκτός and -σχετός 
Comp. ἀν-, ἀντ-, ἀπ-, év-, ἐπ-, KaT-, μετ-, παρ-, περι-, προ-, προσ-, 
προσ-αν-, συν-. ὕπερ-, ὑπ- ; 
On -a forms in imperf. and aor. see §§ 84, 88; on εἴχοσαν ὃ 84, 
; p. 194. 
ζεύγνυμι (11. β. ὃ) -ἐζευξα 
yoke 
Comp. συν- 
ζέω (1. a) 
boil 
VERBAL ἵεστός 
For ζέσω, but inflected like πλέω (only pte.). 
thw (VIL) ἔζησα t ζήσω 
live (The strong ζήσομαι 
aor. ἐβίων, from a cognate stem, was 
used in Attic.) 
Comp. dva-, cuv-. For flexion see ὃ 84. 


Lwypéw Periphr. ἐϊζώγρημαι 
take alive (ἔσῃ ζωγρῶν) 
ζώννυμι (II. β. 6) -ἔζωσα ζώσω -ἔζωσμαι 
σίγα ἐζωσάμην -ζώσομαι 
ἐϊώννυον 


Comp. ἀνα", δια-, περι-, ὕπο- 
For flexion of present stem, and thematic forms in it, see § 86. 
ἥκω (I. α) fg ἥξω 
have come 
ἧκον 
Comp. ἀν-, καθ- 
For perfect endings in the present indic, see § 92. 


240 A GRAMMAR OF NEW TESTAMENT GREEK. [ὃ 95 


ἡσσόω (VII.) ἡσσώθην ἥττημαι 


ENS κα» 
defeat 


ἡττῶμαι 
For variation of oo and rr see ὃ 43; for confusion of classes of 
Contracta § 84. 
θάλλω (VIL) -€arovT . 
bloom 
Comp. dva- 


θάπτω (VIL) ἐτάφην ἔθαψα 
bury 
Comp. συν- 
θέλω (1. a) ἠθέλησα θελήσω 
will 
ἤθελον (ὃ 83 (1)) 


never ἐθέλω 


θεμελιοώ (VIL) ἐθεμελίωσα θεμελιώσω τεθεμελίωμαι 
Sound τεθεμελιώμην 
(onaugm. see 
§ 83 (4)). 
θερίζω (Υ11.) ἐθέρισα θερίσω t 
reap : ἐθερίσθην 
θιγγάνω (11. a) ἔθιγον 
touch 
θλάω (I. a) -θλασθήσομαι 
crush 
Comp. συν- 
θλίβω (I. a) τέθλιμμαι 
press 
θλίβομαι 
-ἔθλιβον 
Compe. ἀπο-, συν- 
«θνήσκω (IV.) -ἔθανον -θανοῦμαι τέθνηκα 
die ἐτεθνήκειν 
-ἔθνησκον Never com- 
VERBAL θνητός pounded : 
Comp. ἀπο-, συντ-απο- see Prol. 
115. int. 
τεθνάναι 


Ac 14! D. 


§ 95] VERBS. _ 241 
θραύω (I. a) τέθραυσμαι 
bruise 
θύω (VIT.) ἔθυσα τέθυμαι 
sacrifice ἐτύθην 
ἔθυον 
θύομαι 
VERBAL θυτός 
-ἴημι (1. β. δ) -εἶναι (not in- -ἧκα -ἥσω -εἶκα (Lk 
send dic.—see  -€yv(see§83 -ἐθήσομαι 1039 C*) 
§ 88, p. 209). (4)) -ἕἔωμαι 
Imperf. see ἀφίημι (indic.) 


VERBAL -ἑτός 


-εἶμαι (pte. ) 


Comp. dv-, ἀφ-, καθ-, Tap-, συν- 
For flexion of present and strong aor. stems see ὃ 86 ff. 


-ἱκνέομαι «ἰκόμην 
(IL. B. a. iii.) 
arrive 


Comp. ἀφ-, δι-, ἐφ- 


ἱλάσκομαι (1Υ.) 
propttiate 


«ἴστημι (1. 8.0) ἔστην 
stand 
-ἱστάμην 
VERBAL -στατός 


ἱλάσθην 
ἔστησα στήσω ἕστηκα 
ἐστάθην στήσομαι (εὐϊστήκειν 
: σταθήσομαι -ἑστώς (ptc.) 
(8 65, 98) 


« Ἄνα. 1 
-“ἐστακαῖ 


Comp. ἀν-, ἀνθ-, ἀντι-καθ- (augm., § 88 (2)), ἀφ-, ἀπο-καθ- (-κατ- 
§ 83 (2)), δι-, ἐν-, ἐξ-, ἐξ-αν-, ἐφ-, ἐπ-αν-, καθ-, κατ-εφ-, μεθ-, 
παρ-; περι-, προ-, συν-, συν-εφ- 

For the flexion of present and strong aorist stems, and thematic 
substitutes occurring .there, see § 86. Present, weak aor. and 
future act. are transitive, as is the new perf. -éoraxa. From 
the intrans. perf. ἕστηκα comes a new verb στήκω stand 
(intr.), and from the present two new verbs ἱστάνω and στάνω 
stand (tr.): cf. the early formation ὀλέκω from ὀλώλεκα, ete. 


καθαίρω (VIT.) 
cleanse 
VERBAL καθαρτός 
Comp. δια-, ἐκ- 


κεκάθαρμαι 


A denominative from καθαρός 


242 A GRAMMAR. OF NEW TESTAMENT GREEK. [§ 95 


καθαρίζω (VII.) éxabépicat καθαριῶ κεκαθέρισμαι 
cleanse ἐκαθερίσθην (-ίσω 1 Jn 19 A 33) 
Comp. δια- 


Forms where augment or reduplication appear show in good MSS 
an ε after 6 (ἃ 2, B 7, A 7, C ἡ, Ὁ 3, L 4), which is certainly 
felt to be a second augment following what popular etymology 
took to be κατά. -It seems better not to allow any phonetic 
cause here, despite Thackeray 74. Such a combination as 


Mk 14? καθαρίσθητι. . . ἐκαθερίσθη seems decisive (see ὃ 33). 
καθέζομαι ( VII.) -καθεσθείς 
sit 
ἐκαθεζόμην 


Comp. παρα- 
The simplex ἕζομαι was extinct, and the compound was not felt 
to be such: cf. κάθημαι. 
καθεύδω (I. a) 
sleep 
ἐκάθευδον 
Another verb with simplex extinct. 


καθίζω ἐκάθισα καθίσω κεκάθικα F 
sit (seat) καθίσομαι 
Comp. ἀνα-, ἐπι-, παρα-, περι-, συγ-. See on καθέζομαι. 
καίω (Υ11.) -ékdny ἔκαυσα καύσω κέκαυμαι 
burn ἐκαύθην καυθήσομαι 
καίομαι -καήσομαι 
«ἔκαιον There is no reason for 
Comp. ék-, κατα- following Veitch in 
parsing καυσούμενα 


(2 Pet 310) asa future. 
Καυσόω is extant in 
medd. and elsewhere. 
The present stem κἄω (not κάω), found often in Attic, is obsolete. 


καλέω (VIT.) ἐκάλεσα καλέσω Τ κέκληκα 

call ἐκαλεσάμην καλέσομαι 7 κέκλημαι 
καλοῦμαι ἐκλήθην κληθήσομαι (ἐ)κεκλήμην 
ἐκάλουν Fut. ἃ. m. is developed 
VERBAL κλητός from the aor. (Att. 
Comp. ἀντι-, εἰσ-, éy-, émi-, peta-, παρα-, καλῶ, -οὔμαι), perhaps 

προ-, προσ-, Tuv=, συμ-παρα- in class, times, 
καλύπτω(Υ 11.) Papyri have ἐκάλυψα καλύψω κεκάλυμμαι 


cover instances of ἐκαλύφθην καλυφθήσομαι 


§ 95] VERBS. 243 


καλύπτομαι ἐκαλύφην. 
VERBAL καλυπτός 
Comp. ἄνα-, ἀπο-, ἐπι-, κατα-, Tapa-, περι-, συγ- 
ἑ καμμύω (Υ11.) ἐκάμμυσα 
shut (eyes) 

For κατα-μύω, from the dialectic form κατ-: it was dissociated 
from κατά by its abnormal form, and its compound character 
forgotten. 

κάμνω (11. a) ἔκαμον 
grow weary 
κάμπτω (Υ 11.) ἔκαμψα κάμψω 
bend 
Comp. ἀνα-, συγ- 
καυχάομαι (Υ 11.) ᾿ἐκαυχησάμην καυχήσομαι κεκαύχημαι 
boast 
Comp. éy-, κατα- 
For 2 sg. pres. καυχᾶσαι see ὃ 85. 
κεῖμαι (1. δ) 
lie 
ἐκείμην 
Compe. ἀνα-, ἄντι-, ἀπο-, ἐπι-, κατα-, παρα-, περι-, προ-, συν-ανα 
For flexion see ὃ 87. 


κείρω (VII.) ἔκειρα 
shear ἐκειράμην 
-«κέλλω (VIT.) “ἔκειλα 
run ashore 
Comp. ἐπι- 


In Ac. 274! ἐπώκειλαν appears in w (against SAB*C 33): this is 
from ὀκέλλω, probably in origin a compound (ὦ = Skt. @-, seen 
in ὠκεανός, ὀδύρομαι etc.). 


κεράννυμι (1. B) ἐκέρασα κεκέρασμαι T 
mur, : κέκρᾶμαι 
VERBAL -κρᾶτός 
Comp. συγ- 
κερδαίνω (Υ11.) ἐκέρδανα κερδανῶ 
gain ἐκέρδησα κερδήσω 
κερδηθήσομαι 


These alternatives occur together even 
in one context (1 Co 9531, where 
read κερδάνω). 

VOL. I]. PART Il.—17 


244 A GRAMMAR OF NEW TESTAMENT GREEK. [§ 95 


κινέω (II. 6) ἐκίνησα κινήσω 
move ἐκινήθην 
κινοῦμαι 


VERBAL -κινητός 

Comp. peta-, συγ- 

The root is xi-: the older non-thematic present κίνυμαι may have 

stood in the quotation adapted in Ac 1778 (see Camb. Bibl. 
Essays 481). 


κίχρημι (I. a. δ) ἔχρησα 
lend 
κλαίω (VIT.) ἔκλαυσα κλαύσω tT 
weep » κλαύσομαι 
ἔκλαιον 
The Attic κλάω (cf. on καίω above) is obsolete. 
kdaw (IIT) ἔκλασα 
break ἐκλάσθην 
κλῶμαι 
Comp. ék-, κατα- 
κλείω (VIT.) -ἔκλεισα 
lock -ἐκλείσθην κλείσω κέκλεισμαι 
CoP. ἀπο-, ἐκ-, κατα-, συγ- 
κλίνω (VII.) ἔκλινα κλϊνῶ κέκλϊκα 
lean ἐκλίθην κλιθήσομαι 
Comp. ἀνα", ἐκ-, κατα-, προσ- 
κομίζω (VII) “ἐκόμισα κομίσομαιΤ 
carry ἐκομισάμην κομιοῦμαι 
-ἐκομιζόμην 
Comp. ἐκ-, συγ- 
κόπτω (Υ11.) -ἐκόπην ἔκοψα κόψω 
beat ἐκοψάμην κόψομαι 
κόπτομαι κοπήσομαι 
ἔκοπτον 
ἐκοπτόμην 
Comp. ἀνα-, ἀπο-, ἐκ-, ἐγ-, κατα-, προ-, προσ- 
κορέννυμι (II. β) ἐκορέσθην κεκόρεσμαι 
satiate 
κράζω (VII) ἔκραγον ἔκραξα κράξω Kexpaya 
ery ἐκέκραξα κεκράξομαι 
ἐκραΐον 
Comp. ἀνα- 


See Prol. 147. 


§ 95] VERBS. 245 


κρέμαμαι (1. δ) ἐκρέμασα 
hang ἐκρεμάσθην 
-ἐκρέμετο 
Comp. ἐκ- 
For the flexion see ὃ 87: the imperfect is thematised. 
κρίνω (VII.) ἔκρινα Kpwv@ KeKpika 
judge ἐκρινάμην κριθήσομαι κεκρίκειν 
κρίνομαι ἐκρίθην κέκριμαι 
ἔκρινον 
ἐκρινόμην 


VERBAL -kpités 
Comp. dva-, dvt-atro-, ἀπο-, δια-, éy-, ἐπι-, κατα-, συγ-, cuv-uTO-, 
ὑπο- 
κρύπτω (Υ1].) ἐκρύβην 7 ἔκρυψα κέκρυμμαι 
-κρύβω f (I. a) 
hide 
~expuBov 
VERBAL κρυπτός 
Comp. dtro-, éy-, περι- 
Present stem only in περιέκρυβεν (Lk 124), which is probably 
impf., and not a newly-formed aorist. 


«κτείνω ( 11.) -ἔκτεινα “κτενῶ 
«κτέννω (11.) «ἐκτάνθην 
«κτέννυμι (11.) 

kall 


Comp. ἀπο-. The simplex is obsolete. See Vocabulary, s.v. 
ἀποκτείνω, for the strange absence of this verb from vernacular 
sources till late. The pres. stem in -νυμι, a modification of the 
classical ἀποκτίνυμι (Kiihner-Blass 469) under the influence of 
the « found in the future, is quotable from Mk 125 B, where 
x*ACD have ἀποκτέννοντες and N° the older ἀποκτιννύντες : 
B is presumably right here, but the -evvo- may be recognised 
in Lk 124 8A, unless D be followed with -evo- —here B adopts 
the normal form and is probably wrong. We may explain 
-κτένω as developed from ἔκτεινα by the proportion ἔμεινα : μένω. 
-Κτέννω is possibly a thematising of κτέννυμι. The welter of 
variants in present stem may perhaps be linked with the 
word’s record, which is not at all clear. See for these forms 
the app. crit. at Mt 108 23’, Mk 12°, Lk 124 1334, 2 Co 36 
and Rev 611, 

κτίζω ( 11.) ἔκτισα ἔκτισμαι 
Sound ἐκτίσθην 


246 A GRAMMAR OF NEW TESTAMENT GREEK.  [§ 95 


-κυέω (VIT.) -ἐκύησα 


-κύω (Ν11.) 
bring forth 
Comp. ἀπο-. The simplex was not obsolete: see Vocabulary, s.v. 
ἀποκυέω. The NT occurrence of the present is not decisive 
between κύω and κυέω: ἐκύομεν in Is 5918 encourages us to 
accent ἀποκύει (with Ti) in Jas 115. 


-κυλίω (VIT.) «ἐκύλισα -κυλίσω -κεκύλισμαι 
roll 
ἐκυλιόμην 
Comp. ἀνα-, ἀπο-, mpoo-. The pres. stem, reformed from the 
aorist (*éxvAwd-ca) as early as Aristophanes, has naturally 
ousted such forms as κυλίνδω (etc.): cf. Brugmann-Thumb Gr. 
360. 


-κυνέω (II. β. a) -ἐκύνησα -κυνήσω 
kiss 
-ἐκύνουν 
Comp. rpoo-. The simplex is obsolete, and the stem-forming 
suffix -ve- has been taken over in other tense-stems, so that 
προσκυνέω has the normal flexion of a contract verb, (Aor. 
ἔκυσα as late as Aristotle.) 


-κύπτω (VIT.) ἔκυψα 
stoop 
Comp. dva-, κατα-, Tapa-, συγ- 


Aayxavw ἔλαχον 
(II. a. a) 
draw by lot 


λακέω (VIL) ἐλάκησα 
burst : 
Blass (on Ac 118) compares διαλᾶκήσασα in Aristoph. Vub. 410, 
as giving a better meaning than λάσκω (aor. ἔλακον, weak aor. 
ἐλἄκησα), to which the NT word is usually referred. 


λαμβάνω ἔλαβον ἐλήμφθην᾽' λήμψομαι εἴληφα 
(II. a. a) (-a, p. 208) -“λημφθή-ὀ εἴλημμαι 
take ἐλαβόμην σομαι ἡ (38g. -εἴληπ- 
λαμβάνομαι ται.) 
ἐλάμβανον 


VERBAL τλμηπτός ἢ 
Comp. ἀνα, ἀντι-, ἀπο-, ἐπι-, κατα-, μετα-, παρα-, προ-, προσ-, 
προσ-ανα-, συλ-, συν-αντι-, συμ-παρα-, συμ-περι-, ὗπο- 


§ 95] VERBS. 247 


The intruded p (from present stem) is firmly established in fut. 
and weak aor., and in the verbal: see Prol. 56. Literary in- 
fluence produced numerous relapses in late MSS: cf. also 
Mt 2122 C, Jn 148 CL 1624 C, 1 Co 38 C, Jas 112 Ὁ. 


λάμπω (I. a) ἔλαμψα λάμψω 
shine 
Comp. ék-, περι- 


λανθάνω ἔλαθον -λέλησμαι 
(II. aa) -ἐλαθόμην 
_ escape notice 
Comp. ἐκ-, ἐπι- 


λέγω (1. α) -ἐλεξάμην 
say -ἐλέχθην 
λέγομαι 
ἔλεγον 
ἐλεγόμην 


VERBAL τλεκτός 
Comp. ἀντι-, δια-, ἐπι-, προ- 
For its suppletives see under εἴρω and ,/Fer. 


λέγω (I. a) -ἔλεξα -λέξω λέλεγμαι 
gather -ἐλεξάμην 
-λέγομαι 
-ἐλεγόμην 


VERBAL “λεκτός 
Ομ». ἐκ-, ἐπι-, κατα-, παρα-, συλ- 


λείπω (I. a) -ἔλιπον ἔλειψα T λείψω λέλειμμαι 
λιμπάνω (II. α. a) ἐλείφην 
leave 
ἔλειπον 
ἐλειπόμην 
ἐλίμπανον 


VERBAL πλειπτός 

Comp. ἀπο-, *Sia-, éy-kata-, ἐκ-, ἐπι-, κατα-, περι-, 
and λιμπάνω). 

There is a good deal of variation in MSS between ἔλειπον and 
ἔλιπον. Since itacism does not produce λίψω or ἐλίφθην, we 
may probably assume that the scribes of our uncials meant 
aorist-stem forms when they wrote Aur-: they may often be 
repeating variants taken from unprofessional copies where 
itacism was really responsible. See WH App.? 162. 


*Gmo- (λείπω 


248 A GRAMMAR OF NEW TESTAMENT GREEK. [ὃ 95 


λούω (I. a) ἔλουσα 
wash ἐλουσάμην 
λούομαι 
Comp. ἀπο- 


μακαρίζω (VIT.) 
congratulate 


μανθάνω ἔμαθον 
(II. a. a) 
learn 


Comp. κατα- 
μαρτύρομαι (VIT.) 
protest 
ἐμαρτυρόμην 
Comp. δια-, προ- 
-μάσσω (Υ 11.) 
wipe 
-ἔμασσον 


-ἔμαξα 


-μάσσομαι 
Comp. ἀπο-, ἐκ- 
μεγαλύνω (Υ11.) 
magnify 
ἐμεγάλυνον 


ἐμεγάλυνα 


ἐμεγαλυνόμην 
μεθύω (Υ11.) 


be drunken 


ἐμεθύσθην 


μεθύσκομαι (IV.) 
get drunk 
μέλλω (111. a. a) 
intend 
ἤμελλον 
On augment see § 83 (1). 


-ἐμελήθην 


ἔμελλον. 
μέλει (1. α) 
impers. 7 matters 
ἔμελε 
-«μέλομαι (I. a) 
care 
-ἐμελόμην 


ἐμαρτῦράμην 


ἐμεγαλύνθην 


λέλουμαι 
Heb 10” 
λελουσμέ- 
vornD*P: 
so Jn 131° 
E, and 
LXxX. 
μακαριῶ 
μεμάθηκα 
μεγαλυνθήσομαι 
μελλήσω 
-μελήσομαι 
-μεληθήσομαι 


§ 95] VERBS. 249 


VERBAL -μελητός 
Comp. ém-, μετα- 
μέμφομαι (I. a) ἐμεμψάμην 
blame (later uncials at Mk 73). 
VERBAL -μεμπτός 
μένω (I, a) ἔμεινα μενῶ μεμένηκα 
remain μεμενήκειν 
ἔμενον 
ComP. dva-, δια-, ἐμ-, ἐπι-, κατα-, παρα-, περι-, προσ-, συμ-παρα-, 
ὕπο- 
μιαίνω (Υ 11.) ἐμιάνθην μεμίαμμαι T 
defile cl, μεμίασμαι 
VERBAL -μιαντός 
-μίγνυμι (11. β. ὁ) ἔμιξα μέμιγμαι 
mix 
-μίγνυμαι 
ComP. συν-ανα- 
Present stem only middle. The correct spelling in pres., weak 
aor. and perf. pass. is we-, but in Hellenistic this is a mere 
matter of orthography. 


μιμνήσκω (ΤΥ...) -ἔμνησα -μνήσω μέμνημαι 
remind ἐμνήσθην μνησθήσομαι 
μιμνήσκομαι é 
ComP. dva-, ἐπ-ανα-, ὑπο- 
μνηστεύω (VII.) ἐμνηστεύθην ἐμνήστευμαι 
betroth Lk 177 peur. 
CD. 
μολύνω (VIT.) ἐμόλῦνα 
defile ἐμολύνθην 
μολύνομαι 
μωραίνω (Υ11.) ἐμώρᾶνα 
make foolish ἐμωράνθην 
«νέμω (I. a) -ἐνεμήθην 
distribute 
Comp. dtro-, δια- 
γήφω (I. a) ἔνηψα T 
be sober 
Comp. dva-, ἐκ- 
vikdw (VIT.) ἐνίκησα νικήσω νενίκηκα 
conquer 


νικῶμαι 


250 A GRAMMAR OF NEW TESTAMENT GREEK. [ὃ 95 


Comp. ὑὕπερ- 
In Rev 217 νικοῦντι appears in AC, but in 27 only A, in 15? C: 
for these confusions οὗ -ἀὼ and -éw stems see ὃ 84, 


νίπτωΐ (VIT.) ἔνιψα 
νίπτομαι T ἐνιψάμην 
wash 


Comp. ἀπο- 
The pres. stem is a back-formation, replacing cl. νίζω, where ζ is 
the regular resultant of y following velar g”. 


vvoow (VII.) -ενύγην ἔνυξα 
pierce 
Comp. κατα- 
νυστάζω (VIT.) ἐνύσταξα 
nod 
Enpaivw (VII) ἐξήρανα ἐξήραμμαι 
dry up ἐξηράνθην Aliso -σμαι 
ξηραίνομαι in earlier 
Greek. 
évpéw (VIT.) ἐξυράμην ξυρήσομαι ἐξύρημαι 
ξύρω (Υ11.) 
shave 


Ξυρασθαι, 1 Co 11°, following κείρασθαι; is not easily construed 
as present (ξυρᾶσθαι, by confusion of -éw and -dw classes), and is 
better taken (ξύρασθαι) as aor. from ξύρω, quotable from Hippo- 
crates and Plutarch: see WH App.? 173 (also above, p. 200). 

-οἴγω, See ἀνοίγω 


οἶδα, see ,ὑεΕειδ 


οἰκοδομέω (Υ11.) ὠκοδόμησα οἰκοδομήσω οἰκοδόμημαι 
build οἰκοδομήθην oikodopn- (Lk6* 8 BL 
φκοδόμουν For the aug- θήσομαι 33 οἷ. ib. 
οἰκοδομοῦμαι ment see 429 T).) 
Comp. ἀν-, ἐπ-, συν- § 83 (6). φὠκοδομήμην 
οἰκτίρω (VII) οἰκτιρήσω ἡ 
pity 


In later authorities generally spelt οἰκτείρω. 
οἴομαι ( 11.) 
think 
The Ist sing. pres. οἶμαι survives—perhaps originally an “ allegro- 


form” of οἴομαι, used in parenthesis (Thumb in Brugmann 
Gr.* 80). 


§ 95] VERBS. 251 
-olxopat (I. a) -ῴχημαι 
have gone 
Comp. παρ- 
«ὀκέλλω, see -κέλλω 
-ὄλλυμι -ὠλόμην -ὥλεσα -ὀλέσω 
(II. B. 6) -dhO 
destroy -ὀλοῦμαι 
-ὄλλυμαι 
-ὠλλύμην 
ComP. ἀπ-, συν-απ-’ 
For fiexion see §§ 86-87, also for thematic forms included. 


-ὅὄλωλα 


ὀμείρομαι (VIT.) 
desire 
“At. Ney. in 1 Th 28 (cf. Job 521, Ps 622 Symm., ὕπερομ. in Irenaeus 
60). Its similarity to the synonymous ἱμείρομαι (denominative 
from ἵμερος) is accidental. It may be a compound of μείρομαι 
obtain, which in the conative present could take the required 
meaning; or we may compare directly the root smer ‘“‘re- 
member.” Debrunner, Jdg. Forsch. xxi. 204, does not con- 
vince me. On the prepositional relic 6- see ὃ 111. 
ὄμνυμι (11. B. Ὁ) ὥμοσα 
swear 
Only one non-thematic form (ὀμνύναι) survives: the present tense 
forms are from ὀμνύω. 


ὁμοιόω (VIT.) ὡμοίωσα ὁμοιώσω -ὡμοίωμαι 
liken ὡμοιώθην ὁμοιοθήσο- -ὁμοίωμαι tT 
Comp. ἀφ- μαι So Heb 73 
CDal.: see 
§ 83 (5). 


ὀνίνημι (11.α.6) ὠνάμην 
profit Only opt. 1 sg. ὀναίμην : see ὃ 88. 


Jor ὠψάμην ὄψομαι 
866 Only subj. ὀφθήσομαι 
ὀπτάνομαι ἦ ὄψησθε Lk 1378, developed from fut. 
ὥφθην 
A suppletive of ὁράω and βλέπω, g.v. 
δράω (VII.) [εἶδον see [ὠψάμην [ὄψομαι ἑόρακα 
see Fed] ὥφθην : see ὀφθήσο- ἑώρακα 
ἑώρων Jér] μαι: see ἑωράκειν 
ὁρῶμαι ν΄ ὁπ) 


-ὁρώμην (ὃ 83 (4)) 


252 A GRAMMAR OF NEW TESTAMENT GREEK. [ὃ 95 


VERBAL ὁρᾶτός 

Comp. ἀφ-, ép-, καθ-, προ- 

A further suppletive in pres. stem active of simplex is βλέπω, 
which is already outstripping ὁρᾶν. The closeness of associa- 
tion between εἶδον and ὁράω is seen in forms like ἀφίδω, ἔφιδε, 
where the aspirate is carried on. 


-ὀργίζω (Υ11.) ὠργίσθην -ὀργιῶ 
anger 
ὀργίζομαι 
Comp. παρ- 
ὀρθόω (Υ11.} -ὥρθωσα -ὀρθώσω 
set upright -«ὠρθώθην Lk 1318 ἀνορθώθη ABDL αἱ. : 
see ὃ 83. 
Comp, dy-, ἐπι-δι- 
ὁρίζω (VIT.) ὥρισα -δριῶ ὥρισμαι 
define ὡρίσθην -ὁδρίσω 
-ὥὡριζον 
Comp. ἀφ-, ἀπο-δι-, προ- 
«ὀρύσσω «ὠρύγην ὥρυξα 
dig (Mt 24% Β -ὠρύχθην 


al., Lk 1239 APQ al.). 
Comp. &-, ἐξ- 
-dtpuvw (VII.) -ὥτρῦνα 
incite 
Comp. παρ- 
ὀφείλω (VII.) ὄφελον (for dpeAov—has become a particle: see Prol. 
owe 200 f.) 
ὦὥφειλον 
ὀφείλομαι 
Comp. προσ- 
παίζω (VII.) -ἔπαιξα f -παίξω 
play «ἐπαίχθην[ἠ΄ = warxOyjooparT 
ἔπαιζον 
Comp. ἐμ- 
παίω (I. a) ἔπαισα 
strike 
πάσχω (IV.) ἔπαθον πέποιθα 
suffer 
VERBAL παθητός 
Comp. προ-, συμ- 


§ 95] VERBS. 253 


πατάσσω (VII.) ἐπάταξα πατάξω 
strike 
A suppletive of τύπτω, ¢.v. 
παύω (I. a) ἔπαυσα -παύσω πέπαυμαι 
stop ἐπαυσάμην παύσομαι 
παύομαι -παήσομαι 7 
ἐπαυόμην 


VERBAL -πα(υ)στός 
Comp. ἀνα-, ἐπ-ανα-, συν-ανα-, κατα- 
᾿Ακατάπαστος is probably an instance of the change οὗ av to a: 
see Prol, 47. 
παχύνω (VIL) ἐπαχύνθην 
Satien 
πείθω (1. a) ἔπεισα πείσω πέποιθα 
counsel ἐπείσθην πεισθήσο- ἐπεποίθειν 
ἔπειθον μαι πέπεισμαι 
πείθομαι 
ἐπειθόμην 
VERBAL πιστός 
Comp. ἀνα- 


πεινάω (Υ 11.) ἐπείνασα πεινάσω 

hunger 

This verb (inflected like τιμάω) has curiously parted from its 
twin διψάω, which keeps its future and aorist stems in -yo, 
although they have both lost the η of pres. stem (διψῆν, 
πεινῆν). Debrunner (Blass 4 43) suggests that association with 
κοπιάω fut. -aow has affected it. In MGr it has pulled over 
διψῶ again (aor. ἐδίψασα). 


-πείρω (VIT.) -ἔέπειρα 
perce 
Comp. περι- 
πέμπω (1. a) ἔπεμψα πέμψω 
send -ἐπεμψάμην 
-ἔπεμπον -ἐπέμφθην 
πέμπομαι 


Comp. ἀνα-, ἐκ-, μετα-, προ-, συμ- 
The punctiliar character of the root is in keeping with the rarity 
of its occurrence in the present stem. 


πετάννυμι (II. β. 6) -ἐπέτασα 
spread 
Comp. ἐκ- 


254 A QRAMMAR OF NEW TESTAMENT GREEK. [ὃ 95 


πήγνυμι (11. Bb) ἔπηξα 
Fasten 
Comp. προσ- 
-πιπλάω (VIT.) ἔπλησα πλησθήσο- πέπλησμαι 
jill ἐπλήσθην μαι 
Comp. ἐμ- 
The simplex present πίμπλημι does not occur in NT. 
πίμπρημι (1. δ) -ἔπρησα 
burn 
πίμπραμαι 
Comp. ἐμ- (πιπρᾶσθαι) : so Ac 28° x* (Bw πίμπρασθαι) 
+ maw (VII.) ἐπίασα 
seize ἐπιάσθην 
πιέζω (Υ11.) πεπίεσμαι 
press 


Like ἀμφιάσαι against ἀμφιέσαι, πιάσαι is said to be Doric: see 
Thumb in Brugmann Gr. 78. Differentiated meaning seems to 
have set in, for πιάσαι (cf. MGr πιάνω) has the new force seize. 


πικραίνω (VII.) ἐπίκρᾶνα πικρανῶ 
embitter ἐπικράνθην 
πικραίνομαι 
Comp. παρα- 
πίνω (11. a) ἔπιον ἐπόθην πίομαι (δ δῦ, πέπωκα 
drink — Inf. πεῖν see also Prol. 184 n.°). 
ἔπινον (§ 38). 


VERBAL (ποτός) whence noun πότος 
Comp. κατα-, συμ- 
πιπράσκω (1V.) ἐπράθην πέπρᾶκα 
sell 
ἐπίπρασκον πέπραμαι 
πιπράσκομαι 
πίπτω (1. ἃ) ἔπεσον πεσοῦμαι πέπτωκα 
fall (-a, § 88) 
ἔπιπτον 
Comp. ἀνα-, ἀντι-, ἀπο-, ἐκ-, ἐμ-, ἐπι-, κατα-, παρα-, περι-, προσ-, 


συμ- 
πλάσσω (Υ 11.) ἔπλασα 
shape ἐπλάσθην 


VERBAL πλαστός 


πλατύνω (Υ11.) ἐπλατύνθην πεπλάτυμμαι 
broaden 


§ 95] VERBS. 255 


πλέκω (1. a) -ἐπλάκην ἔπλεξα 
entwine 
-πλέκομαι 
Comp. ἐμ- 


πλέω (1. α) ἔπλευσα 
saal 
ἔπλεον 
Comp. ἀπο-, δια-, ἐκ-, κατα-, παρα-, ὗπο- 
For conjugation of present stem see § 84. 


-πλήσσω ἐπλήγην -ἔπληξα 
(VII.) -ἐπλἄγην 
strike 
-πλήσσομαι 
-ἐπλησσόμην 
Comp. ἐκ-, ἐπι- 
Ac 1312 ἐκπληττόμενος B 33: on this see ὃ 43. The simplex 
passive aor. was used in Attic as suppletive for τύπτω : it was 
distinguished by its 7 from the form used in compound. 


πλύνω (VIL) ἔπλῦνα 
wash (clothes) 
ἔπλυνον 
Comp. ἀπο- 


πνέω (I. a) ἔπνευσα 
breathe 
VERBAL -πνευστός 
Comp. éx-, ἐμ-, ὕπο- 
For flexion of present stem see § 84. 
-πνίγω (1. a) -ἐπνΐγην ἔπνιξα 
choke 
-πνίγομαι 
ἔπνιγον 
ἐπνιγόμην 
VERBAL πνικτός 
Comp. ἀπο-, συμ- 


ποιέω ( VII.) ἐποίησα ποιήσω πεποίηκα 
make ἐποιησάμην ποιήσομαι πεποιήκειν 
ἐποίουν πεποίημαι 
ποιοῦμαι 
ἐποιούμην 


VERBAL -ποιητός 
Comp. περι-, προσ- 


256 A GRAMMAR OF NEW TESTAMENT GREEK. [ὃ 95 


The spellings without ἢ are rare: thus δὲ in Lk 114? ποῖσαι, but 
the same MS has ποιῆσαι normaliy. In earlier Greek and in 
papyri ποῶ is found, and papyri have forms without y See 
p. 73. 

ποιμαίνω (VIT.) ἐποίμανα T ποιμανῶ 
shepherd 
πράσσω (VII.) ἔπραξα πράξω πέπρᾶχα 
practise πέπραγμαι 
πρίω (I. α) ἐπρίσθην 
saw 
-ἐπριόμην 
Comp. δια- 
προφητεύω (Υ11.) ἐπροφήτευσα προφητεύσω 
prophesy Later MSS augment προεφ. : see ὃ 83 (2), (7). 
ἐπροφήτευον 
πταίω (1. α) ἔπταισα 
stumble 
VERBAL -πταιστός 
πτύσσω (VII.) ἔπτυξα 
roll up 
Comp. ἀνα- 
-πτύω (VII.) ἔπτυσα -πτύσω 
spit -πτυσθήσομαι 
“ἐπτυον 
Comp. ἐκ-, ἐμ- 
πυνθάνομαι ἐπυθόμην 
(II. a. a) 
ascertain 
ἐπυνθανόμην 
-ῥαίνω (Υ 11.) -ρέραμμαι 
sprinkle 

Comp. περι- 

The simplex may be original at Rev 1918 (pf. pte. pass.): see 
WH App.? 140. On the breathing see § 40. 


partite (VIT.) ἐράντισα ῥεράντισμαι 
sprinkle ἐραντισάμην (On breathing 
see p. 100). 
ῥαπίζω (VII.) ἐράπισα 
beat 


ῥάσσω (VII.) 
strike 


§ 95] VERBS. 257 


An independent verb (only in Mk 918 D 565), on which see 
Thumb in Brugmann G'r.* 346. 
pew (I. a) (p)unv pevow 
flow subj. -pu@ 
Come. παρα- 


ῥήγνυμι (11. β. δ) ἔ(ρ)ρηξα ῥήζω 
(β 83 (9)) 
ῥήσσω (VII.) 
break 
ῥήγνυμαι 
-ἐρησσόμην 
Comp. δια-, περι-, προσ- 
ῥιζόω (VIL) -ἐρίζωσα -ριζωθήσο- ἐρρίζωμαι 
root Only subj. μαι 
-ριζώσητε 
«ἐριζώθην 
Comp. ἐκ- 
ῥιπτέω (VII.) épupa tT ἔρ(ρ)ιμμαι 
cast see § 83 (8). 


Comp. ἀπο-, ἐπι- (both with single p following) 

The pres. ῥιπτέω is said to differ from firrw as iacto from iacio 
(μετὰ σφοδρότητος pirtw): the latter does not occur in NT, 
the former once. 


ῥύομαι (VIT.) ἐ(ργρυσάμην ῥύσομαι 

deliver (§ 83 (9)) 

ἐρύσθην 
ῥώννυμι ἔρρωμαι 
(II. 8. ὁ) 

strengthen 
σαλπίζω (Ν 11.) ἐσάλπισαΤ σαλπίσω t 

trumpet 


This is an instance of a new aorist and future coming from 
present (caAmtyy-yw—> -ίζω phonetically): cl. σαλπίγξω, ete. 


σβέννυμι (11. B. Ὁ) ἔσβεσα σβέσω 
quench 
σβέννυμαι 
VERBAL -σβεστός 
Note phonetic spelling ζβέννυτε in 1 Th 51° B*D,*FG. ΑἹ] the 


present stem forms are still non-thematic. 


258 A GRAMMAR OF NEW TESTAMENT GREEK. [8 95 


σέβομαι (1. a) ] ἐσεβάσθην 
worship 
VERBAL σεβαστός 
In earlier Greek separate, but in Hellenistic the derivative verb 
is used as a suppletive, σέβομαι having no aorist in continued 
use. 


-ceiw (1. a) ἔσεισα σείσω 
shake ἐσείσθην 
σείομαι 
Comp. ἀνα-, δια-, κατα- 
σημαίνω (Υ 11.) ἐσήμανα T 
signify 
ἐσήμαινον 
one (I. a) σέσηπα 
rot 
σκάπτω (VIT.) ἔσκαψα -ἔσκαμμαι 


Comp. κατα- 


-σκέπτομαι (VII.) -ἐσκεψάμην σ-σκέψομαι 
watch 

Comp. ém- 

In Attic σκοπέω (only pres. stem in NT) was used as suppletive 
for the present stem. We find ἐπισκοποῦντες in Heb 12% in 
this sense, and in some papyri and inscriptions. It is perhaps 
significant that when interpolated (by A and the late MSS) 
in 1 Pet 5?, it is a denominative from ἐπίσκοπος, and an 


entirely separate word. 


σκληρύνω (VIT.) ἐσκλήρῦνα 
harden , ἐσκληρύνθην 
ἐσκληρυνόμην 
σκύλλω (Υ11.) ἔσκυλμαι 
distress 
σκύλλομαι 
σπαράσσω (VII.) ἐσπάραξα 
convulse 
Comp. συν- 
σπάω (111.) -ἔσπασα -σπάσω -ἔσπασμαι 
draw ἐσπασάμην 
-ἐσπώμην -ἐσπάσθην 


ΜΈΕΒΑΙ, -σπαστός 
ΟοΜ». ἀνα-, ἀπο-, δια-, ἐπι-, περι- 


§ 95] VERBS. 259 


σπείρω (VII.) ἐσπάρην ἔσπειρα ἔσπαρμαι 
sow 
ἔσπειρον 
σπείρομαι 
Comp. δια-, ἐπι- 
σπεύδω (I. a) ἔσπευσα 
hasten 
ἔσπευδον 


σπουδάζω (VIT.) ἐσπούδασα σπουδάσω T 
be zealous (cl. -opac) 


-στέλλω (VIT.) -ἐστάλην -ἔστειλα -στελῶ -ἔσταλκα 
send -ἐστειλάμην -ἔσταλμαι 
-ἔστελλον 
στέλλομαι 
-ἐστελλόμην 
Comp. ἀπο-, δια-, ἐξτ-απο-, ἐπι-, κατα-, συν-, συν-απο, ὗπο- 
στενάζω (VIT.) ἐστέναξα 
groan 
Comp. dva-, συ- 
Ἱ στήκω (I. a) 
stand 
ἔστηκον 
A new present (MGr στέκω) formed from ἕστηκα---866 ἵστημι. 


στηρίζω (VIT.) ἐστήριξα στηρίξω ἐστήριγμαι 
establish ἐστήρισα (LK στηρίσω 7 (2 Th 95 B) 
VERBAL -στηρικτός 95! BCE al, Acv15% CH Lk 2282 
Comp. ém- SABL αἱ.). 
ἐστηρίχθην 
στρέφω (I. a) ἐστράφην ἔστρεψα -στρέψω ἔστραμμαι 
twin -“στραφήσο- ἔστρεμμα: ὕ: 
-ἔστρεφον μαι see WH 
στρέφομαι App.7177f. 
Comp. ἀνα-, ἅπο-, δια-, ék-, ἐπι-, κατα-, μετα-, συ-, ὗπο- 
στρώννυμι (II. β. δ) ἔστρωσα ἔστρωμαι 
strew -ἐστρώθην 
ἐστρώννυον 


VERBAL -στρωτός 
ComP. κατα-, ὗπο- 
σφάΐζω (VII) ἐσφάγην ἔσφαξα σφάξω ἔσφαγμαι 
slay 
Comp. κατα- 
VOL. II. PART I1.—18 


260 A GRAMMAR OF 


-σχηματίζω (Ν 11.) 
Fashion 
-σχηματίζομαι 
ΟΜ». μετα-, συ- 
σχίζω (Υ11.) 
split 
Comp. δια- 
owtw (VII) 
save 
σῴζομαι 
ἐσῳϊόμην 
Comp. δια-, ἐκ- 


NEW TESTAMENT GREEK. [Ἃ8 95 


-ἐσχημάτισα -σχηματίσω 


ἔσχισα σχίσω 

ἐσχίσθην 

ἔσωσα σώσω σέσωκα 

ἐσώθην σωθήσομαι σέσωί(σ)μαι 

Ac 49- σται 

BDEP,— 
ται δίᾳ, 
Eph 28,— 
-μένος P,— 
-opevosrest. 


The « subser. is best confined to pres.: see K.Bl. 1. 544, and 


above, § 36, d (a). 


ταράσσω (Υ 11.) 
disturb 
ἐτάρασσον 
ταράσσομαι 
Comp. δια-, ἐκ- 


τάσσω (Υ 11.) -ἐτάγην 


arrange 
τάσσομαι 
VERBAL τακτός 


ἐτάραξα τετάραγμαι 
ἐταράχθην 

ἔταξα -τάξομαι τέταχα 
ἐταξάμην -“ταγήσομαι τέταγμαι 
ἐτάχθην 


Comp. ἀνα-, ἀντι-, ἀπο-, δια-, ἐν-, ἐπι-, ἐπι-δια-, mpo-, προσ-, συν-, 


ὕπο- 
-τείνω (Ὑ11.} 
stretch 
-ἔτεινον 
-τείνομαι 


“ἔτεινα -τενῶ 


ComP. ἐκ-, ἐπ-εκ-, παρα-, προ-, ὕπερ-εκ--. Simplex obsolete. 


τελέω (VII.) 
finish 


τελοῦμαι 


ComP. ἀπο-, δια-, ἐκ-, ἐπι-, συν- 


-τέλλω (Υ 11.) 
accomplish ? 


Comp. ἀνα-, ἐξ-ανα-, ἐν- 


ἐτέλεσα -τελέσω τετέλεκα 

ἐτελέσθην τελεσθήσο- τετέλεσμαι 
μαι 

-ἔτειλα -τελοῦμαι -τέταλκα 

-ἐτειλάμην -τέταλμαι 


§ 95] VERBS. 261 


The simplex occurs very rarely in early poetry. Its central 
meaning is quite doubtful: it is very probably a conflation of 
originally distinct roots. 

-τέμνω (11. a) -ἔτεμον «ἐτμήθην -τέτμημαι 
cut 

VERBAL -τμητός 

Comp. περι-, συν- 


τήκω (1. a) τακήσομαι 
melt So 2 Pet 515 Ὁ, rest 
τήκομαι mostly τήκεται. 


See WH App.? 103, where Hort suggests τήξεται (quotable from 
one MS of Hippocrates) to account for the variants: Mayor 
seems to approve the emendation. 


τίθημι (I. 8) θεῖναι ete. (in- ἔθηκα θήσω τέθεικα T 
set dic. wanting ἐτέθην -θήσομαι τέθειμαι 
in NT). -teOjoopar -ἐτεθείμην 
ἐτίθουν ἐθέμην 

τίθεμαι The cl. -τέ- 
-ἐτιθέμην θηκα oc- 
VERBAL -θετός cursin Βα Ὁ 
Comp. ἀνα-, ἀντι-δια-, ἀπο-, δια-, ἐκ-, ἐπι-, κατα-, ΠΡ 95.595 
μετα-, παρα-, περι-, προ-, προσ-, προσ-ανα-, (111 A.D.), 
συν-, συν-επι-, συγ-κατα,- ὗπο- but the 
For the flexion see ὃ 86 f. other form, 
formed 
after ἀφ- 


etka, had 
long been 
current. 
tiktw(I.a.a@) ἔτεκον ἐτέχθην τέξομαι 
bear (child) 
-τινάσσω (VIT.) -ἐτίναξα 
shake -ἐτιναξάμην 
Comp. ἀπο-, ἐκ- 
τίνω (II. a) τείσω 
pay The cl. orthography— 
Comp. ἀπο- usually written τίσω. 
-τρέπω (I. a) -ἐτράπην -ἔτρεψα -τραπήσομαι 
turn -ἐτρεψάμην 
“τρέπομαι 
-ἐτρεπόμην 
Comp. ἀνα-, ἀπο-, ἐκ-, ἐν-, ἐπι-, μετα-, περι-, προ- 


262 A GRAMMAR OF NEW TESTAMENT GREEK. [8 95 


τρέφω (1. a) -ἐτράφην ἔθρεψα τέθραμμαι 
nurture -ἐθρεψάμην 
τρέφομαι 


Comp. ἀνα-, ἐν-, ἐκ- 

The root is θρεῴ, which loses its first aspiration when the second 
is present. 

τρέχω (I. a) [ἔδραμον] 
run 
ἔτρεχον 

ComP. εἰσ-, ἐπι-συν-, κατα-, TEpt=, προ-, προσ-, συν-, ὕπο- 

The suppletive is from ,/dpeux: cf. δρόμος. The root is θρεχ, 
whence future θρέξομαι. 


-τρίβω (I. a) -ἔτριψα «τρίψω -τέτριμμαι 
rub -τρίβήσομαι 
-ἔτριβον 
Comp. δια-, συν- 
τυγχάνω (II. a) ἔτυχον τέτυχα tf Heb 
happen 86 x*AD* 
Comp. év-, ἐπι-, παρα-, συν-, ὕπερ-εν τέτευχα 2b. B 
τετύχηκα UW, 
P 33 
-τυλίσσω (VIL) -ἐτύλιξα -τετύλιγμαι 
wind 
Comp. év- 
τύπτω (VIL) [ἐπλήγην] [ἔπαισα] [πατάξω] 
strike [ἐπάταξα] 
ἔτυπτον 
τύπτομαι 


For the suppletives see παίω, πλήσσω and πατάσσω: cf. for the 
cl. usage the excursus on ‘The Defective Verb τύπτω in 


Sandys and Paley’s Demosthenes, vol. 11, pp. 207-211. 
ὑφαίνω (Υ11.) 


weave 
VERBAL ὕφαντός 
J/pay ἔφαγον φάγομαι ἡ 
eat 
Suppletive to ἐσθίω, q.v. 
φαίνω (VII) ἐφάνην ἐφᾶνα φανοῦμαι 
shine φανήσομαι 
φαίνομαι In Rev 8:32 18% φάνῃ is best thus 
VERBAL -φαντός accented, to mean “shine” rather 


Comp. dva-, ἐπι- than ‘‘appear” (φανῃ). 


§ 95] VERBS. 263 


7 -φαύσκω (LV.) : -φαύσω 
T φώσκω (LV.) 
Comp. ἐπι- 
dawn 
-ἔφωσκον 
The relations of these words are not quite clear. ᾿Επιφαύσει 
might be the future of (ἐπιλφάξω (dae “shone” in Homer— 
still surviving in Aratus). The Homeric πιφαύσκω = “ declare.” 
᾿Επιφώσκω can be quoted from P Lond. 130° (=i. p. 134), a 


os fees 


φέρω (1. a) ἤνεγκον (not ἤνεγκα οἴσω -ἐνήνοχα 
bear indic.) ἠνέχθην 
ἔφερον 
φέρομαι 
ἐφερόμην 


Compe. ἀνα-, ἀπο-, δια-, εἰσ-, ἐκ-, ἐπι-, κατα-, παρα-, παρ-εισ-, 
περι-, προ-, προσ-, συμ-, ὗπο- 
On the question οὗ aorist Aktionsart in ἔφερον see Prol.? 129, 
247. 
φείδομαι (1. a) ἐφεισάμην φείσομαι 
spare 
φεύγω (1. a) ἔφυγον : φεύξομαι πέφευγα 
flee 
Compe. ἀπο-, δια-, ἐκ-, κατα- 
φημί (I. 6) 
say 
ἔφην 
φάσκω (1.) 
ἔφασκον 
Comp. σύμ- 
Of φημί only φημί, φησί, φασί, and ἔφη are found. 
φθάνω (II. a) ἔφθασα ἐφθακα 
anticipate 
Comp. προ- 
φθέγγομαι (I. a) ἐφθεγξάμην 
utter 
Comp. ἀπο- 
φθείρω (VII.) ἐφθάρην ἔφθειρα φθερῶ -ἔφθαρμαι 
corrupt φθαρήσομαι 
ἔφθειρον 
φθείρομαι 


264 <A GRAMMAR OF 


VERBAL φθαρτός ᾿ 
Comp. δια-, κατα- 
φορέω (Υ11.) 
wear 
VERBAL -φορητός 
φράσσω (VII.) ἐφράγην 
shut 
φρυάσσομαι (VII.) 
neigh 
Middle in cl. Gr. 


φυλάσσω (VII.) 
guard 
ἐφύλασσον 
φυλάσσομαι 
Comp. δια- 
φύω (Υ11.), ἐφύηνΤ 
grow 
VERBAL -φυτός 
Comp. ék-, συμ- 


NEW TESTAMENT GREEK. [ὃ 95 


ἐφόρεσα 


ἔφραξα 


ἐφρύαξα ἡ 


ἐφύλαξα 
ἐφυλαξάμην 


φορέσω 


φραγήσομαι 


φυλάξω 


In Mt 2453-- Mk 13°8 we may accent éx@vy pres. (transitive) or 


ἐκφυῇ aor. (intrans.). 


Since this late strong aor. accounts for 


every other occurrence of φύω or compounds in NT, except 
φύουσα in Heb 1215, there is a presumption in favour of it 
here, as it enables us to take both verbs as aorist. To parse 
ἐκφύῃ as aor. subj. act. is certainly wrong, for ἔφυν was obsolete 
and ἐφύην took its place. 


φωτίζω (VII) 
illuminate 
χαίρω (VIT.) ἐχάρην 
rejouce 
ἔχαιρον 
Comp. συγ- 


χαλάω (Υ11.) 
let down 

χαρίζομαι (VIT.) 
grant 


ἐφώτισα 


ἐφωτίσθην 


ἐχαρισάμην 
ἐχαρίσθην ἡ 


φωτίσω  πεφώτισμαι 

φωτιῶ (Rev 235 δὲ) 

χαρήσομαι 

Χαροῦσιν in Rev 110 ΤῊ 
is perhaps invented 
out of vg gaudebunt. 
Ἔπιχαρεῖ can be quoted 
from an Alexandrian 
Preisigke Sammelbuch 
no. 1323. 

Xadaow 


χαρίσομαι tT κεχάρισμαι 
χαρισθήσομαι ἦ 


§ 95] VERBS. 265 


-xéw (I. a) -ἔχεα -χεῶ T -κέχυμαι ᾿ 
-χύννω ἡ (IT. a) -ἐχύθην -χυθήσομαι 
pour There is difficulty as to the parsing of ἐκχέετε in Rev 
-ἔχεον 161 and συνέχεον Ac 2157, That the active indic, 
-ἔχυννον present is nowhere χέω in NT proves nothing for 
-χέομαι impf., where there was not the same confusion with 
-χύννομαι the future ; and Luke is not likely to have altered 
-ἐχυννόμην aor. συνέχεαν so as to suggest an unintended impf. 


Moreover, he does use ἐπιχέων. In Rev the appear- 
ance of an uncontracted form ἐκχέετε (imper.) might, 
as WH App.? 172 suggest, be aor. with termination 
of strong aor. But the syntactical argument they 
use is not very strong; and the open forms cited 
from LXX by WS 115 n.% justify us equally in 
parsing this form as (iterative) present, considering 
the ways of the author of Rev. 


Comp. ék-, ἐπι-, κατα-, cuy-, ὕπερ-εκ- 


χρήομαι (Υ 11.) ἐχρησάμην κέχρημαι 
use 
ἐχρώμην 
VERBAL χρηστός 
Comp. kata-, συγ- 
For flexion see ὃ 85: the ordinary -ἄω flexion has considerably 


affected it. 
χρηματίζω (VIT.) ἐχρημάτισα χρηματίσω ἡ 
be called 
χρηματίζω (VIT.) ἐχρηματίσθην κεχρημάτισ- 
warn μαι 


? 


Two entirely distinct words, the former from χρήματα “ business’ 
—cf. our phrase “trading as X. & Co.”; the latter from an 
equivalent of χρησμός “ oracle.” 


xptw (VII.) ἔχρῖσα 
anoint -ἐχρισάμην (Rev 518, accenting ἔγχρισαι: 
VERBAL χρῖστός οἵ, vg inunge). 
Comp. éy-, ἐπι- 
χρονίζω (VIT.) xpoviowt (Heb 10278*D*) 
tarry χρονιῶ (2b, Aw). 
χωρίζω (VIT.) ἐχώρισα χωρίσω κεχώρισμαι 
separate ἐχωρίσθην 
χωρίζομαι 


Comp. ἀπο-, δια- 


266 A GRAMMAR OF NEW TESTAMENT GREEK. 


ψάλλω (VIL) ψαλῶ 
sing 
ψεύδομαι (1. a) ἐψευσάμην 
lie 
«ψύχω (1. a) -“ἔψυξα ψυύγήσομαι F 
cool 
Comp. ἀνα-, ἀπο-, ἐκ-, κατα- 
-ὠθέω (Υ11.} -ὦσα + 
thrust -έωσα (Ac. 748 X*E) 
-ὠθοῦμαι -ὠσάμην T 


Comp. ἀπ- ἐξ- 


PRINTED BY MORRISON AND GIBB LTD., EDINSURGH 


[ὃ 95 


PAD DTT: 
WORD-FORMATION. 


§ 96. Our subject hitherto has been exclusively the form of 
words as isolated units, prepared to take their place in the 
combinations which make up speech. But we are not yet 
ready to pass on to Syntax, the study of principles governing 
these combinations: we have still to investigate important 
features in the history of the words themselves. We must go 
a step further back, and examine words before they are 
prepared for their place in a sentence by receiving inflexions 
of case or number, of person or mood. The larger part of this 
field must be excluded from our present survey. The Science 
of Language, or Grammar in the older and wider sense, is 
bound to concern itself with Semantics or Semasiology, the 
scientific investigation of the development of the meaning of 
words. In the same way it is bound to pass beyond Syntax, 
and study the phenomena classed under Rhetoric or Stylistic. 

§ 97. But convenience demands an artificial limiting of 
what is involved in “Grammar”; and as we limit Syntax 
by the period or sentence, so we shall limit the study of single 
words within the provinces of Accidence and Word-formation. 
This latter division, upon which we now enter, will include 
the elements from which Semantics must start. 

§ 98. There are two main divisions in the province before us, 
which we will take in retrograde order. First comes, then, the 
formation of Compounds, the combination of separate words 
within the history of the Greek language. Then follows the 
study of formative elements within words—comparative research 
into the function of prefixes and suffixes which ceased to be 
separate words (assuming that they ever were such) in a pre- 
historic stage of Indo-European speech. Under both these 
headings we shall have to deal separately with Nouns and 
Verbs. 

VOL. 11. PART III.—19 


268 A GRAMMAR OF NEW TESTAMENT GREEK. [8 99-101 


WORD-COMPOSITION. 
NOUNS. 


§ 99. There is a marked difference between the languages of 
our family, and even between those which are closely related, in 
the extent to which noun-composition is employed. It is one of 
the most obvious differences between Greek and Latin, between 
German and English—the last pair fellow-members of one 
branch in the Indo-European family. English has less objection 
to compounds than Latin had ; but we are familiar enough with 
the instinct that bids us coin our compounds in Greek rather 
than with native elements. We mentioned Semasiology just 
now. German philologists use die Bedeutungslehre, “ meaning- 
science,” where we invent a Greek combination and transfer 
it to our own tongue. 

§ 100. The extreme example of proclivity towards com- 
pound-making is found in Sanskrit, where all sorts of relations 
are expressed by fitting words together, leaving their syn- 
tactical functions to be inferred. The Indian grammarians 
classified compounds with their usual acuteness, and our 
scientific accounts of them to-day are largely based upon 
the work of these early philologists. Macdonell (Sanskrit 
Grammar, p. 155) gives a good illustration: “ Kalidasa 
describes a river as ‘ wave-agitation-loquacious-bird-row-girdle- 
stringed,’ while we should say ‘her girdle-string is a row of 
birds loquacious because of the agitation of the waves.’” It 
will be noticed that Sanskrit has the advantage in brevity, but 
not in clearness. Greek, with its characteristic instinct of 
proportion, avoids the overdoing of this practice, especially in 
prose ; and Hellenistic, as we might expect, restricts it some- 
what more. Colloquial speech is not prone to multiply cumbrous 
words, and the more complex compounds are naturally avoided 
in the artless prose of everyday life. 

§ 101. But it is possible to go too far in suspecting special 
culture when compounds are freely found. Within limits, a 
compound may be an actually simpler locution than its 
equivalent with the elements separated ; and MGr vernacular 
shows that the genius of the language in respect of this resource 
has not changed. If we find Paul using a compound which 
we cannot trace elsewhere, we may provisionally allow him 


§ 101-103] WORD-COMPOSITION—NOUNS. 269 


to have coined it without mferring any literary qualities in a 
man who can avail himself of such means of expressing a new 
thought. A modern peasant could do the same. 

§ 102. For the general principles of classi- 
fication and the large literature thereon, 
reference may be made to Brugmann Grundriss? τι. 1. 35-40 and 
49-120. Α scientifically complete classification can only be 
made on comparative lines, and will not be attempted here, 
as we are not going beyond the compounds that occur in the 
NT. For our purpose we cannot do better than to apply 
with a few variations the classification devised by the Indian 
grammarians. We may quote Macdonell’s summary (Sanskrit 
Grammar, p. 155): “ The most convenient division is into the 
three classes of Co-ordinatives, Determinatives, and Possessives. 
The Determinatives, so-called because the former member deter- 
mines (or qualifies) the latter, are of two kinds, Dependent and 
Descriptive. Possessives are secondary compounds, consisting 
chiefly of Determinatives turned into adjectives.” The Sanskrit 
names for the classes, which will be given below, are with one 
exception typical examples of the class thus described. 

§ 103. I. Co-ordinative Compounds (Skt. 
Dvandva, 1.6. “ couple’’) unite two or more 
words of the same class: thus Skt. aho-ratram “day and 
night.”” Numerals afford plentiful examples : δώδεκα, duodecim, 
δεκαοκτώ, eighteen, twenty-one, etc., may be compared with the 
forms in which and occurs, as the classical tpevoxaidexa or 
our three and twentieth. Apart from these there are very few 
examples of Dvandva in Greek: the adverb ? νυχθήμερον “ by 
night and day” is the solitary NT instance.* ᾿Αρτόκρεας, 


Classification. 


Co-ordinatives. 


1 (Mr. E. E. Genner cites πλουθυγίεια health and wealth. (Aristoph.”’.)—ED. ] 

2(So J. H. M. here: but see below, p. 283. “Substantivised neut. from 
*yyxOnuepos,” Brugmann Gird.* τι. i. 99. Generally taken as direct obj. of 
πεποίηκα in 2 Co 11*°.—Ep. ] 

3The vox nihili δευτερόπρωτος might be said to belong to this class as 
probably as to any other: we may record it here simply because this is the 
first class named. Plummer’s note on Lk 6! gives the various attempts made 
to explain it. It is an early Western interpolation. WH App.* 58 (q.v.) 
accept Meyer’s explanation ofits origin; Burkitt (Gospel Hist. 81 n.) supposes a 
dittography caBBaTwBatw@, from which “ Batu was erroneously expanded 
into Sevreporpwrw.” Cod. e reads sabbato mane=év σαββάτῳ πρωί: cf. Mrs. 
Lewis’s conjecture of πρωΐ for πρῶτον in Jn 1*!, from a new reading of syr*!® 
(Hap T xx. 229 ff.). 


270 A GRAMMAR OF NEW TESTAMENT GREEK. [§ 103-104 


found in Persius 6°° as a borrowed word, is taken by Brugmann 
to mean “‘ bread and meat”: it may be Hellenistic, or a survival 
from classical vocabulary. Brugmann notes that Dvandva 
is commoner in MGr: thus γυναικόπαιδα “women and 
children,” μαχαιροπέρονα “‘ knives and forks.” It even 
appears in verbs, as aveBoxateBaivw “ pace up and down ” 
(Thumb Hdb. 112). The papyri give us χορτάρακον (P Lond 
1171%) (=u1. p. 178) (8 B.c.), which Grenfell and Hunt take 
as “hay and aracus,” comparing κριθόπυρος (P Petr i. p. 78) 
“barley and wheat.” (But see Mayser 468 on this. Kenyon 
takes it as “‘aracus for fodder.”) Mayser 469 adds ναυκλη- 
ρομάχιμος. We may fairly assume that the usage was known 
in Hellenistic. 

§ 104. II. Closely akin to these compounds 
are the iteratives, products of the reduplicating 
tendency which forms so many words in our own vernacular, 
and takes a large part in the history of the Indo-European 
verb-system. Unlike Sanskrit and Latin, Greek hardly ever 
fuses the two forms into one word, but the principle is not 
really affected. An exception is πέρπερος (whence περπερεύομαι), 
which is not attested till a later date than Latin perperam is 
found: it may quite possibly be borrowed from perperus. It 
comes from the pronominal pero- “other” (see below, p. 279, 
under πέρυσι), and the iteration produces elative force, ‘‘ other 
than what should be” ; ef. uses of érepos. The reduplication 
produces distributive force in δύο δύο “ two by two ” (see Proleg. 
97, and add P Oxy vi. p. 310 note), πρασιαὶ mpactat, συμπόσια 
συμπόσια, δεσμὰς δεσμάς : cf. Skt. ékackas “each one,” yddyad 
“what each time,’ Lat. qwisqus, Avestan nmdne-nmdne 
“in each house.” A modern Indian dialect like Gujarati may 
complete the case against Semitism! Emphasis of various 
kinds may be seen in πλέον πλέον, μᾶλλον μᾶλλον, magis 
magis, ““ more and more”; μέγας péyas! (Proleg. l.c.) “ very 
great’; ὅσον ὅσον (Aristophanes, LXX, Lk 5? Ὁ, Heb 102?) 
“only just so much as.” The last is like our idiomatic phrase 
“only so-so”’: English iteratives like “tut tut!”, ‘ hear 


Iteratives. 


1 Archiv v. 163 cites an Egyptian inser. of 29 B.c. with dedication θεῶι 
μεγάλωι μεγάλωι ὑψίστωι: O. Rubensohn regards it as referring to Jehovah, 
though it may or may not be from a Jew. The collocation μέγας μέγας was 
however specially Egyptian. 


§ 104-105] WORD-COMPOSITION—-NOUNS. 271 


hear!”, “come, come!”, “a red, red rose,” will supply 
further illustration. See Delbriick, Grd. v. (iii.) 139 ff. 
§105. III. Dependent Determinatives 
D ao nonee (Skt. Tatpurusa “his man,” abbreviated T.P.) 
eterminatives. 
form a very large class. In these the first 
element stands in a case relationship to the second. As in 
other compounds, the immense majority of instances show the 
bare stem in the first part. There are a few exceptions, of which 
we may name the following as NT words: νουνεχής (adv. in 
Mk 124) acc., δδοιπόρος (noun and verb in NT) loc., ἄριστον 
(Ξε ἀψέρι -" στόν, loc. of obsolete noun=morning—cf. ἦρι 
Eng. ere—with p.p.p. of ν' ed to eat), Διόσκουροι (sons of Heaven) 
Ac 281; gen. A few specimens may be given to show these 
compounds in the various case relations : it should be premised 
that the selection of a case is often ambiguous. 
Accusative.—@coceBys God-worshipping, γεωργός (=yn- 
opyos) earth-working, tiller, φωσφόρος light-bringing, daystar, 
. κακοῦργος (=Kakoepyos) evil-doer (hence by analogy παν- 
ovpyos one who will do anything, knave). 
Genitive —sarpiapyns ruler of a clan (πατριά), αἱματεκ- 
χυσία shedding of blood ; οἰκοδεσπότης lord of a house. 
Ablative.—évorrer7s fallen from heaven, πατροπαράδοτος 
handed down from our fathers, ποδαπός (see Prol. 95, Hellen. 
ποταπός) coming from whence ? 
Dative.—avOpwradpecxos man-pleaser, εἰδωλόθυτος  sacri- 
ficed to rdols. 
Locative.—yeiuappos flowing in the winter, γονυπετής 
falling on the knees, ὀφθαλμόδουλος serving under the eyes. 
Instrumental.—otapodopntos river-borne, χειρόγραφον 
written with the hand, αἰχμάλωτος spear-captured, κιθαρῳδός 
(Ξεκιθαρα-αοιδός) singing with the harp, θεοδίδακτος God-taught. 


There are a great many compounds of this form which will easily be 
placed in one of the categories described above. Some words follow on 
which special notes are desirable. (As throughout this chapter, the 
original compound is given whenever it exists: in many cases the NT 
only shows a secondary derivative.) 

᾿Αλεκτοροφωνία is noted by Blass (Gr. 68) as “* peculiar, there being 
no conceivable adjective from which it can be derived.” It is no doubt 
genitivally dependent, cock-crowing, with its second element modelled on 
συμφωνία, ete. It is vulgar, as Blass observes. 


272 A GRAMMAR OF NEW TESTAMENT GREEK. _ [ 105 


᾿Αλλοτριεπίσκοπος (1 Pet 41°, vg alienorum appetitor) may very well 
be a coined word: of course such a statement only means that we have 
no other occurrences at present, and that the coining of such a compound 
is entirely according to usage in Greek of every kind. The elision of its 

is determined by 8B and other authorities: for the ἀλλοτριοεπίσκοπος 
of the a-text MSS we might compare ἱπποἴΐατρος P Oxy i. 92 (iv/a.D.), 
Ramsay C. and B. ii. 662 (no. 627), δειγματοάρτ(ης) and χωματοεπι- 
μ(ελήτης) in P Lond 1159. (=iii. p. 113) (Hermopolis, a.p. 145-7), the 
former also in P Oxy i. 638 (ii/iili A.p.). Probably the retention of the 
o was normal in freshly coined words at a rather later period than that of 
the NT Books: WM 124 ἢ describes the other as “‘ more correct.” See 
further above, § 32. 

‘Appodoyos may be tentatively given as the basis of (cuv)appodoyew, 
though its authority (see LS) is nil, and the verb may be simply formed 
from analogy. It would mean joint-gathering (accus. dep.), with the 
verb to fit together. 

Αὐτός forms sundry compounds that call for comment. In two of 
them the airos is nominative, so that the words fall into the next class. 
In three it is instrumental, viz. αὐθαίρετος self-chosen, αὐτόματος self- 
thought, spontaneous (*-pards p.p.p. of men to think), αὐτοκατάκριτος 
self-condemned. Probably attédwpos is the same, self-detected: its 
usage (see Jebb on Soph. Ant. 51) depends on φωράω rather than on dap. 
The dative appears in αὐτάρκης self-sufficient, content: the vernacular 
use of the word practically lets the αὐτός go—see Vocabulary s.v. Also 
dative is αὐθάδης (airo-Fadys, from ν᾿ sudd of ἥδομαι, ἁνδάνω, suavis 
etc.) self-pleasing, reckless. 

Βατταλογέω (Mt 67 \B—WH App.” 158) presumes an (unquotable) 
*Barradoyos, best perhaps taken as from ἔβατταλο-λόγος by haplology. 
It is argued in Vocabulary s.v. that Barrados, the nickname of Demos- 
thenes, may have meant gabbler : in that case Aramaic battdl (Wellhausen) 
may be only accidentally similar. βατταρίζω stammer is another possible 
source, for *Sarrapo-Aoyéo might suffer haplology after assimilation. 
The reading of D (gr.) βλαττολ. is akin to Latin blatero (i.e. *mlatero, 
Eng. blether; the root perhaps in Skt. mleccha “ foreign-speaking ’’). 
The uncertainty of the word’s history makes it doubtful whether Class IV. 
or V. should not claim it. 

Βλάσφημος probably shows the reduced form of the stem of βλάβος, 
thus injury-speaking (accus. dep.): see further Vocabulary s.v. 

Γαξοφυλάκιον (Strabo p. 319) is derived by Blass (Gr. 68) from 
γαζοφύλαξ treasure-warden (gen. dep.). 

Γλωσσόκομον (Jn 12° 1339) is what holds (κομίζει) the mouthpiece of a 
flute (yAérra—cef. Blass ib.). Phrynichus (Rutherford 181) tells us how 
the degenerate epigoni sacrificed the “ correct”? Attic γλωττοκομεῖον, 
and widened its use so as to describe ἃ case for ‘‘ books, clothes, money, 
or anything else.” See Vocabulary s.v. for papyrus citations for the 
Attic form (still in W), and the shortened Hellenistic form. 

Δεξιολάβος, a d.d. in Ac 2373 (vg lanceartt), is supposed to mean 


§ 106] WORD-COMPOSITION—NOUNS. 273 


taking (a spear) in the right hand (instr. or loc. dependence). In military 
phraseology the spear was always connected with the right, as the shield 
with the left. It was certainly not a coined word, but as it does not re- 
appear till vii/A.p. we must suppose it a technical term of limited range. 

Anpnyopos public orator (whence -péw) starts best perhaps from δῆμος 
ἀγείρειν contionem facere (so accus. dep.), the connotation of a harangue 
coming from the conditions familiar from Homer down. 

Εἰλικρινής (on breathing see WH App.? 151) is as yet unsolved. 
That the second element is from κρίνω seems clear, so that the meaning 
is —tested, —discerned, sincere. But neither the Homeric εἰλίποδες 
(βόες) nor Theocritus’ εἱλιτενής, an epithet of a grass, seems to help us 
for the first element, and it must be left for the present. 

Since εὖ is the neuter of an old adjective (Homeric js, οἵ. Skt. vasu 
good, Zd. vanhu, O. Pers. Daéraya-vahu (Darius) =possessing goods), some 
of its cpds. may as well come here as in the other classes. Thus εὐεργέτης 
benefactor might be bonorum factor. But it is better to divide them be- 
tween Classes IV. and V. 

Θεοστυγής (see SH on Rom 139) has sometimes been taken as accusa- 
tivally dependent, God-hating; but there seems no reason for deserting 
the ancient evidence for God-hated (dative). Similarly the proper name 
Θεόφιλος means God’s beloved, dear to God (gen. or dat.): the other 
meaning would be expressed by μισόθεος, φιλύθεος respectively. (As 
a Jewish name Theophilus appears in papyri: it carries on such OT 
names as Jedidiah.) 

Θυμομαχέω --Ἐθυμομάχος is not found—seems to be instr. dep., to 
fight with zest or heat, so to quarrel hotly. So λογομάχος fighting with 
words (**‘ speaking daggers ’’). 

Oupwpds=Oupa-Fopds door-guardian (ὁράω, Eng. ware) has gen. 
dependence. Cf. xnzo-Fopés, κηπουρός, gardener; otkoupds house- 
guardian (Tit 25 a-text). The 1st decl. nouns properly produced com- 
pounds in -wpds, as πυλωρός gate-guardian, τιμωρός honour-guardian, 
avenger; but analogy sometimes produced in vernacular documents 
forms like θυρουρός. 

Κενεμβατεύω ““ according to a probably certain conjecture, Col 218" 
(Blass, Gr. 67) ἀέρα κενεμβατεύων for ἃ ἑόρακεν ἐμβ. The verb ἐμβατεύω 
is good vernacular, and, in the process of copying, might easily take the 
place of a 4.\., which would probably be a coinage of Paul’s own: a con- 
jecture developed by such a succession of great scholars 1 is tempting. 
C. Taylor, to whom the conjecture in its final form is due, suggested that 
the phrase was based on the Rabbinic “‘ fly in the air with nothing to rest 
on.” Field’s rather doctrinaire objection to a verb not found elsewhere, 
and not formed according to classical rules—which are far from “‘ inviol- 
able’ in Hellenistic—would be easily turned by reading (with Peake) 
κενεμβατῶν, an existing word, and assuming that the common ἐμβατεύειν 


1 See its history in J. R. Harris’s Side-lights on NT Research (1909), 
pp. 198 f. 


274 A GRAMMAR OF NEW TESTAMENT GREEK. [8 105 


ousted a rather rare expression. The new verb will mean to invade the 
void (accus. dep.), and with the internal accus. ἀέρα to tread the empty air. 

*Kapadoxos may perhaps be assumed as basis of (ἀπο)καραδοκία 
(see p. 299), but its formation is not at all clear. If δέκομαι (Att. 
δέχομαι) originally meant to stretch out (cf. Brugmann Grd. τι. 465), this 
(hypothetical) adjective head-stretching would supply a good basis for 
καραδοκέω to expect eagerly. See further on the noun (possibly a coinage 
of Paul—the verb is in Polybius) Vocabulary s.v. 

Ναυᾶγός (vais and βάγνυμι 2) must be interpreted on the same lines 
as its Latin synonym naufragus: accus. dependence is perhaps simplest, 
one who has wrecked his ship. Ναύκληρος is complex, as it combines 
ναύκληρος with a ship as his portion (a compound of Class V. below) with 
a dissimilated form of ναύκρᾶρος ship-master, from *xpaopos head 
(κρᾶσ-πεδον, κρανίον, Lat. cerebrum for ceresrom): there are semasio- 
logical difficulties here—see LS. 

Oikoupyds (ἅ.λ.), the true reading in Tit 25, is a compound with 
locative dependence, home-worker, or it may be exactly parallel with 
γεωργός, Which has accus. 

᾿᾽Ορθοτομέω (2 Ti 215) occurs in Pr 3° 115, of levelling or straightening 
a road—* cutting straight the path of Truth,” for the pilgrims’ progress 
thereon, would be an attractive meaning. But it is simpler to compare 
(with Grimm) καινοτομεῖν to innovate, where the second element has 
faded: ὀρθ. will then be to direct, apply faithfully, as men speak of “a 
straight talk.” So practically vg recte tractantem, and RV. 

Πατρολώας and μητρολώας in 1 Ti 19 are spelt with o (not a) by the best 
MSS: see WH App.? 159. The classical πατραλοίας seems to make father- 
thrasher (ado(t)dm) the meaning.? The levelling tendency of analogy has 
caused all the r nouns to substitute po in compounds for the pa (i.e. r) 
which was originally found (cf. Skt. pitr-, Goth. bropru-): here the 
tendency has even affected words in which the a belongs to the second 
element of the compound. This is simpler than Radermacher’s sugges- 
tion (Gr.’ 35, 737) that ὄλλυμι was in mind. 

Πλεονέκτης should apparently mean one who has more (accus. dep.). 
But from the first it is one who would have more. It is difficult to see 
where the desiderative element came in: is the -έκτης really *é£-rys ? 
But cf. αἰσχροκερδής below, p. 284. 

Πλήμμυρα may be placed here tentatively, on the assumption that the 
pp (See ὃ 41) only represents a popular connexion with πλήν, like πλημμελής. 


*[This “ probably certain conjecture’ has been gravely discounted by 
Ramsay's discovery of the word ἐμβατεύω in inserr. from Klaros as a t.t. of 
the Mysteries. His essay ‘‘ The Mysteries in their Relation to St. Paul” 
(Contemp. Review, Aug. 1913, pp. 198 ff.) is republished in The Teaching of 
Paul, pp. 287 ff.). So now Vocab. 206a.—Ep.] 

* The ἃ 15 taken by Brugmann (@rd.? τι. i. 92) as due to analogy of nouns that 
lengthened the vowel at the junction through contraction, as orparayés=sirto- 
ago-s. 

* For ὦ or w replacing οἱ see p. 83. [WH spell both words with w.—Eb.] 


§ 105] WORD-COMPOSITION—NOUNS. 275 


Boisacq regards πλη- as a gradation-doublet of πλω (πλώω, E. flood): 
μύρομαι, orig. to flood, supplies the second part. In that case the first 
element would be a primitive noun, accusativally or instrumentally 
dependent. 

Πληροφορέω, which has no *zAnpoddpos behind it, may come from the 
analogy of cpds. like τελεσφορέω (from τελεσφόρος maturity-bringing). 
The verb starts from to bring in full (accus. dep.). Its meanings in ver- 
nacular Greek may be seen in Deissmann LAF 82 f. (5386 f.), Milligan on 
1 Th 1°: see also Lightfoot on Col 4”. 

Ποδήρης must mean reaching to the feet (dat. dep.), but its second 
element is not clear. The root of ἀραρίσκω is generally given, and may 
be right, though feet-fitting is not quite what we want. The -npys in 
ἀμφήρης, ἐπήρης, τριήρης al., is connected with the root of our row, 
rudder, and is less suitable still. 

Προσωπολήμπτης is one of the few genuine examples of “᾿ Jewish 
Greek.” It is a natural Greek coinage from the phrase πρόσωπον λαμβάνειν 
(accus. dep.), which was a literal rendering of 0°35 δ). ἃ notion not 
provided with a real equivalent in Greek. Prol. 13 gives an estimate of 
the place of such locutions in the language. The compound (with -pta 
and -πτέω) serves as a good illustration of the fact that word-composi- 
tion was a resource of language still very much alive in Hellenistic. 

Σκύβαλον (Ph 38) was a vernacular word, found in papyri. Wetstein 
gives a mass of passages from Κοινή literature, and one or two from earlier 
writers. The derivation from oxwp is quite impossible phonetically. 
The alternative (based on Suidas), that it is a contraction of és κύνας 
βαλεῖν, aS σκορακίζω is formed from ἐς κόρακας, is not to be vetoed so 
decisively: it might be a popular abbreviation, in form anticipating 
partly the MGr στόν for ἐς τόν, etc. In that case we have a compound 
with quasi-dative dependence. Popular association with σκώρ might 
account for the predominance of the meaning stercus. 

Σπερμολόγος picking up seeds (applied to a bird in Aristophanes and 
other writers) has accus. dependence : for the development of its meaning, 
see comms. on Ac 178, 

Συκοφάντης (whence “τέω) fig-showing (accus. dep.) seems clear in 
its composition, but the meaning of the metaphor which thus describes 
an informer is still uncertain. 

φρεναπάτης mind-deceiving (accus. dep.), conceited. Blass (Gr. 68) 
notes its occurrence in P Grenf i. 11: 19 (ii/B.c.), a literary text. 

Χειροτόνος (hence “€w) stretching the hand (Aischylus) has accus. 
dependence. 

Χορηγός (χορός -- ἀγός) chorus-bringing or -leading, originally used 
mostly of the equipment of a chorus (a λειτουργία at Athens), was early 
generalised to mean one who supplies the cost for anything. The irregular 
n (following p) is probably due to the analogy of other compounds in 
-ηγός, rather than to Ionic influence (see p. 68). (Cf. Brugmann AVG 304, 
307, Grd.? τι. i. 92.) In one of its meanings ἀρχηγός may belong here : 
see below, p. 277. 


276 A GRAMMAR OF NEW TESTAMENT GREEK. _ [§ 106 


§ 106. IV. We come next to Descriptive 
Determinatives (Skt. Karma-dharaya, a term 
of which the exact application is not certain 
—abbreviation K.D.). In these the first element is a noun, 
adjective, or adverb which describes the second element, 
standing to it in a predicate relation. We classify them 
naturally according to the character of the first element : 


Descriptive 
Determinatives. 


(a) Noun or adjective! or pronoun :—Atrénrys_self-seer, 
κωμόπολις village-town (a πόλιες that is little more than a κώμη--- 
Strabo, Mk 138), μεσότοιχος middle wall, barrier, ἀγριέλαιος 
wild olive, ὁλόκαυτος burnt whole. So in English midsummer, 
ill-will. 

(b) A subdivision of (a) is required for compounds with 
numerals, like our fortnight: the Indian grammarians made a 
special class for these, called Duigu (two-cow). In the nature 
of things these, if they belong to Class IV., can only be plural 
(except possible compounds of one, like the English t.t. single- 
tax: μονογενής only born might be classed here). Thus we 
have Latin decemviri, centumvirt. Out of these arose a natural 
singular, not capable of analysis except by reference to its 
plural: thus decemvir=one of a board of ten. This is found 
in the NT τετραάρχης, one of four rulers. Compare the title 
δεκάπρωτος, found in papyri=one of ten πρῶτοι (cf. Ac 28°), 
δεκανός (whence our dean). 


(c) A large class has an adverb as first element. I+ may 
be the negative a-, av- (y-, Lat. m-, Eng. un-, Skt. a-, an-: 
the reduced form of né): so ἄγνωστος unknown, ἄσοφος unwise, 
etc. The second element might be a noun (like our unfaith) : 
thus in Sophocles apytep=one who is no mother. But these 
compounds have in Hellenistic all become possessive (Class V.) : 
andtwp—father unknown,? with no recorded father (Heb 7°). 
Εὖ forms a good many compounds, as_ εὐεργέτης well-doer, 
benefactor ; so δυσ-, as δυσβάστακτος hardly borne (cf. our mis- 
trust), etc. Many prepositional cpds. belong to this class (IV.c), 


1 An interesting observation is made by Brugmann AVG 362 as to fem. 
cpds. with -o- at the juncture, like ἀκρόπολις. He regards these as dating 
from a time when the adj. did not take a fem. termination to suit grammatical 
gender. 

2 So in the papyri, for illegitimates : see Vocabulary s.v. 


§ 106] WORD-COMPOSITION—NOUNS. 2h 


though Classes V. and VII. are responsible for the major part 
of them: thus πρόσωπον what is towards the eyes, face, προπάτωρ 
forefather, ἐπ-ενδύτης overcoat. We will reserve prepositional 
compounds until we can classify them together. (See §§ 110- 
129.) 


We proceed to comment on special cases : 


Class (a). ᾿Ακροβυστία is not completely explained. ᾿Ακροποσθία, a 
normal Descriptive epd. from ἄκρος and πόσθη with a fresh suffix, is found 
in Hippocrates, and is obviously the original of the LXX word. When a 
word containing a vox obscena was taken from medical vocabulary into 
popular religious speech, it was natural to disguise it: a rare word 
βύστρα-- βύσμα may supply the model. 

᾿Ακρογωνιαῖος presumes a (non-existent) dxpoywvia extreme corner. 

᾿Ακροθίνια, based on ἀκρόθινα (Pindar), top heaps=top of a heap 
(cf. Latin swummus mons), chief spoils. 

᾿Αρεοπαγίτης depends formally on a cpd. ᾿Αρε(ι)όπαγος, which 
never had any real existence: it was at a late period formed afresh 
from ᾿Αρεοπαγίτης (for which cf. Meisterh.* 61, also 43). Lobeck Phryn. 
599 ff. successfully shows that in this and similar words the separate 
words were more in accord with Attic taste than the cpd.; but he totally 
fails in his attempt to eject the Descriptive altogether. His quotations 
illustrate that a locution containing two or three words was fused when a 
derived cpd. was wanted: cf. καλοκἀγαθία and other cpds., while καλο- 
κἀγαθός never occurs till very late (ii/A.D.). On this see Brugmann KVG 
306, who compares terrimotium from terre motus, Ger. langweilig from 
lange Weile etc. So we make the compound Free-Churchmanship out 
of the separate Free Church. 

᾿Αρχηγός (cf ἀρχάγγελος, ἀρχιερεύς) may have as its first element 
ἀρχι: the εἐ is elided before -ἄγός (dyw)—for the ἡ see above under χορηγός 
(p. 275). The alternative ἀρχε- (see Cl. VI.) has probably affected this 
prefix, and when it precedes vowels we cannot tell which of the two to 
presume, in the case of early words: ἀρχι- monopolised the late forma- 
tions. Since there are two distinct functions for the prefix, according as 
it means chief (Cl. IV.) or ruling (Cl. VI.), it is reasonable to conjec- 
ture ἀρχι- as original in the adjectival and dpye- in the verbal function, 
with some natural mixture resulting. How far ἀρχι- is primitive we can 
hardly say. It was shown by Caland and Wackernagel (see the latter’s 
Altind. Gramm. ii. 59 ff.) that from I.E. times -i- replaced the adjective 
suffix -ro- in the first part of cpds., and perhaps replaced other suffixes. 
Cf. καλλι- epds. (as καλλιέλαιος) from καλός. In that case ἀρχι- will 
be either a primitive base-form, or made by analogy from ap xo-, the base 
of the noun dpyés. See Mayser 81. The great majority of the late ἀρχι- 


1[Mr. E. E. Genner points out that the new LS gives ἄκροθις as a subst. 
(like ἀκρόπολις) from a iv/B.c. inser. See Riisch, Gr. d. delph. Inschriften, 
i. 216: ἀκρόθις Erstlingsgabe: τὰν ἀκρόθινα D*’.—Ep. ] 


278 A GRAMMAR OF NEW TESTAMENT GREEK.  [§ 106 


cpds. belong to Cl. IV. The old word ἀρχηγός shows fluctuation in 
meaning between originator and leader, according as the force of 
ἄρχεσθαι and ἄρχειν respectively predominates. The former may really 
go back to ἀρχή -ἰ ἀγος, beginning-leader, a cpd. of Cl. III.: the latter is 
Cl. IV., supreme leader. See Vocabulary s.v. 


Αὐθέντης (whence αὐθεντέω be master of, govern) is a contracted form 
of the Sophoclean αὐτοέντης (=atro-evrns from root of ἀνύειν, Skt. sanott, 
Lat. sons, our sin). The verb is branded as vulgar by Atticists, and is 
accordingly good vernacular, in the same sense as in 1 Tim 912, Latin and 
English join in giving the verb the nuance of our word perpetrate. The 
curious meaning “* murderer ”’ in classical writers comes from an entirely 
different word, derived from αὐτο-θέντης : see Vocabulary 8.0. It is 
classed with atromtys as a epd. 


Anptoupyos, from δαμιοξεργός, is as old as Homer. It seems to 
start from public worker, which developed in two different directions 
into (1) craftsman, (2) magistrate (in some Doric cities). Under (1) the 
idea of skill grew stronger—perhaps from the contrast of a publicly 
recognised workman, contrasted with one who only works for himself— 
and it becomes in philosophy a name for the Creator: cf. Philo De Somn. 
13 fin. (p. 632 M.), where it is contrasted with the higher title κτίστης. 
It is natural to think that the author of Heb 11!° was at least semi- 
conscious of this. 

Ἱεράπολις and Νεάπολις are printed divisim by WH: in Col 41° the 
MSS cannot help us, in Ac 16! CD* are quotable (with the mob) for the 
epd. form. Blass’s “‘ sic divisim antiquiores” is a useless remark, since 
Luke was not an “ antiquior.’ On the whole, however, the oldest 
uncials may be allowed to decide, though the tendency to fuse these 
phrases into cpds. was strong by this time. Μεγαλύπολις, for example, 
occurs in Strabo (i/B.c.). Ramsay (C. and B. ii. p. 681) shows that 
Hieropolis was the local name, and a sign of imperfect Hellenisation. 

[Καλοδιδάσκαλος in Tit 2%, being a ἅ.λ., might be taken as noble 
teacher ; but this would probably be ἕκαλλιδιδάσκαλος, and the other 
cpds. of διδάσκαλος belong to Cl. 111.] 


Κράσπεδον is an old cpd., apparently of this class, from the stem 
to which κέρας, κάρα and κρά(σ)νιον belong: utmost edge seems the 
meaning. 

Λειτουργός, from ληϊτο-εργός (Blass Gr. 8), is later in its attestation 
than its formal derivatives (-€o and -ta). Δήιτον occurs in Herodotus 
(vil. 197) for a set of public offices: this connotation would suggest a 
clerk in government service as the oldest meaning. It is curious to note 
how far it has diverged from the very similar δημιουργός. 

Μεσημβρία, from μέσος and an ablaut form of ἦμαρ, ἡμέρα, answers 
to our mid-day. 

Μεσουράνημα, mid-heaven, in Rey 8:5, is from the verb peroupavéw to 
culminate, of heavenly bodies crossing the meridian, but the presumable 
base Ἐμεσούρανος is not quotable, and may never have existed. 


§ 106] WORD-COMPOSITION—NOUNS. 279 


Νεομηνία new-moon, or rather the day of the new moon, is a derivative 
rather than a direct Descriptive cpd. 

Nedgutos new plant, or newly planted, might equally well be put 
with (c) below. 

Navyyupts, like some other cpds. of πᾶς, loses the r of the stem by the 
influence of the neuter πᾶν, but keeps the original a. General assembly 
represents the meaning. 

Πανοικεί is the locative of an unused ἕπανοικός, which would mean 
whole house: cf. πανστρατιᾷ, mavorxia. Πανοικεί comes under the 
ban of the Atticists (Lobeck Phryn. 514). No doubt it was formed on the 
analogy of other locatives in -εἰ (ἀμισθεί ete.), by substituting οἶκος for 
oixia; nor need we suppose that such formations were ever used in the 
nom. or other cases, being called into being only for the adverb. 

Πέρυσι is another locative without other cases in use, bub it is of 
prehistoric antiquity. Its first element is the primitive pronoun pero- = 
other, Skt. para- : cf. πέραν, perendie, perperam, Eng. far etc. : the second 
is the locative of wet (cf. Féros, vetus)=year, with weak grade. The Skt. 
parut has the same meaning, and only differs by the absence of the 
locative suffix -1. 

Πρῶτος forms a good many cpds. of this class, as in NT πρωτοκαθεδρία, 
πρωτοκλισία, πρωτοστάτης. Πρωτότοκος first-born forms the further 
noun πρωτοτόκια (pl.). Πρωτοτύκος -- bearing a firstborn is an older 
or Syll.8 102418 ὗν ἐνκύμονα πρωτοτόκον). Isidore of Pelusium (Lightfoot 
Col. p. 150) proposed to use this for the crux in Col 1, qs. “* the first 
author of all creation” ; but his admission that he was innovating must 
be noted, and Lightfoot’s exegesis may be maintained—-see also Peake 
EGT in loc. 

Σαρδόνυξ, χρυσόλιθος, χρυσόπρασος and the cpds. in ψευδο- 
(exc. two of Cl. V.) are words of this class that need no comment. 

Σήμερον is an adverb (accus.) like πέρυσι. Its termination is an adj. 
suffix; and the o- represents a pronoun κιο- “ this,’ compared by 
Brugmann KVG 401 with O.E. hiu-diga (Ger. Heute) “‘ to-day.” Hence 
*conpepov, Attic ἔττήμερον, whence the initial consonant was reduced. 

[Σκληροκαρδία might belong to this class, if=‘‘ hard heart.’ It is 
better taken as Possessive (Cl. V.) in origin, =hard-heartedness. | 

Συροφοινίκισσα in Mk 77° NALA stands against Sipa Φοινίκισσα 
Bw and Σύρα Φοίνισσα W al.—the last probably an Atticist correction. 
Συροφοῖνιξ occurs in Lucian (cf. also Juvenal)—see Grimm-Thayer— 
and is the natural antithesis to Λιβυφοῖνιξ. 

φθινοπωρινός (Jude 15: see J. B. Mayor’s excursus, pp. 55 ff.) 
is a time-adjective (p. 358 below) from φθινόπωρον, which is a double 
epd. of this class. “Omdpa =late season (ὥρα orig. =spring, but generalised): 
its first part is the ὁπ- of ὀψέ, ὀπίσω, ὄπισθεν, the association with which 
has ejected the @ that would be regular. In usage this season opened 
as early as July; and a new word was needed to describe the autumn 
proper. The adj. φθινάς and cpds. in φθινο- show that the present- 


280 A GRAMMAR OF NEW TESTAMENT GREEK. — [§ 106 


stem ᾧθιν- was accompanied by an adj. base (perhaps merely analogical) 
containing the v. The season is therefore ὀπώρα φθίνουσα (like μηνὸς 
φθίνοντος, of the days following the 20th). So ἰσημερία ἡ φθινοπωρινή 
(Aristotle) =the autumnal equinox. Pomifer autumnus finds these trees 
ἄκαρπα. 

Χαλκολίβανος, ἁ.λ. in Rev 118 and 2:8, is explained by Hort as “" brass- 
like \i8avos,” i.e. amber, and so “ the glowing metal named from amber 
by this name as well as by ἤλεκτρον. ΤῈ this view of the word be taken, 
it is a noun of the Descriptive class. See Hort’s note, also Swete’s in loc. 

Ψευδής forms ψευδάδελφος, ψευδαπόστολος, ψευδόμαρτυρ, ψευδοπρο- 
φήτης, ψευδόχριστος. 

Class (0). On the whole it seems best to transfer to the next class all 
the remaining nouns under this head, as being essentially adjectival. 
Thus δίδραχμον is (an amount) of two drachme ; διετές is the neuter of 
an adj.—lasting two years; ἡμίωρον is (a space) of half an hour. The 
only exception then will be the rerpadpyns type described above. 

Class (c). Descriptive compounds in which the first element is an 
adverb—a term which of course includes prepositions—need not be 
catalogued in full. The prepositional cpds. will be reserved for the next 
chapter, so that we may bring together all the combinations in which the 
several prepositions are concerned. First come the cpds. made with 
the negative prefix. In the great majority of cases this was from I.E. 
times n, the weak grade of né, which in Greek became d-, as ἄ-γνω(σ)τος 
=i(n)-gnotus, unknown; ἄκων (a-Féxov)=un-willing; or av- before 
vowels, as av-4:0s=un-worthy. That né in the high grade could he 
compounded with nouns, adjectives and participles is proved by such 
formations as ne-fas, némo (ne-hemo), our naught (=not one wight or whit— 
see Skeat H.D.). Even verbs were once capable of taking this prefix : 
there is no essential difference between ne-scio, ndlo (ne-volo) and οὐκ 
οἶδα, ov θέλω, except that the former have become one word. A restric- 
tion of » to nouns, adjectives and participles had established itself in 
LE. times; d-80vatos, im-potens, un-witting, are normal; but we do 
not find ἀ-δύναμαι, im-possum, or the like. In Greek this restriction 
excluded even participles when they still belonged to a verb system 
- ἄκων is found, but not ἀδυνάμενος. (The verbal in -ros does not 
originally belong to the verb system proper: see Prol. 221 f.) One 
class of words still maintains (according to Brugmann KVG 310) the old 
né, viz. words where there is @, é, 0 in Anlaut, which sometimes took né 
instead of n. Thus νήκεστος from νε-ἀκεσ- (ἀκέομαι) etc.: from these 
forms, with initial νη- (va-) as the result of contraction, came such words 
as νή-πιος (=-7F-vos, cf. νη-πύτιος) 5. unintelligent, infant. This account is 
not wholly satisfactory as an explanation of the prefix va- or vy-, but 
no other is forthcoming that will do as well, unless we revert to the theory 


1 There are rare exceptions, like ἀπιστεύσας in P Oxy ii. 237 γ'ὃ (A.D. 186). 
2The old etym. νη- - βέπος word is impossible, as an uncontracted form 
must have occurred in Homer. 


§ 106] oe WORD-COMPOSITION—NOUNS. 281 


of the long vowel n producing va. Among the Cl. IV. cpds. with d- we 
need only name one or two. 

"Ayvoua may be from ἀγνώς, if we may assume this to combine an 
earlier *dyvwds with the stem in -7- (gen. dyv@ros): ἀγνοέω presumes 
the same. 

᾿Αδημονέω is a problem imperfectly solved. T. W. Allen (CR xx. 5) 
takes it from ἀδήμων (Hippocrates, 1 cod.), the negative of a presumed 
δήμων prudent, traceable in proper names, as Dor. Δάμων, familiar in 
pastoral poetry. This he takes from δέδαε, that is / δασ (perh. from 
dems—Skt. damsistha), *Sacpwv becoming δάμων regularly. We might 
take the original meaning of ἀδήμων to be bewildered: cf. the association 
0 ἀδημονῶν and ἀπορῶν in Plato (see LS). 

*Adpatos, not ἀν-, because of the F in Fopdw (cf. Eng. ware). 

"“Aowtos not salvable, past saving, dissolute: on this connotation of 
the -ros verbal see Proleg. 222. 

Other cpds. may be taken as they come. ᾿Αγανακτέω is not completely 
explained, but it may possibly depend on an (unused) Descriptive *ayav- 
dxtns greatly angry: ἄγαν is supposed to be a reduced form of μέγας (cf. 
ingens, both thus from mgyt), and the second element I should tentatively 
regard as an agent noun from Vv ay (ἄχομαι) grieve. 

Διχοτόμος (whence διχοτομέω cut in two) is from δίχα a-two: its 
derivatives are from the active paroxytone, but the adj. appears earlier 
as διχύτομος passive =“* hewyne in to” (Chevy Chase). 

Δυσ- cpds. are primitive : in Skt. we have not only the prefix dus but 
also the noun dosa, harm. Cf. also Gothic prefix tuz-. Of doubtful 
history is δύσκολος (morose, in NT simply difficult). Osthoff (IF iv. 281), 
on βουκόλος, makes probable the existence of an I.E. / gel (with pure 
velar) “‘ to keep, tend,’ found with an extension in our hold. The 
synonymous q“el of Latin colo would serve equally well, since εὔκολος 
(with vu preceding the labiovelar—see Giles Manual? 132) would deter- 
mine the form of its correlate (*S5vcmodos otherwise). The meaning seen 
in colo (e.g. patronum or deum) would suit the Greek very well: δύσκολος 
would be one who ἄνθρωπον οὐκ ἐντρέπεται, So disagreeable. 

Over against these δυσ- compounds stands the more numerous family of 
εὖ, This adverb performs the same function as the Skt. su-, but it does 
not seem possible to equate them. It is assigned by the best authorities 
(see Boisacq Lex. 298 f.) to a primitive esu-s, which makes Gaulish Hsogen(us) 
(Welsh Owen1)=Evyévios. Su- appears in b-yiys, from ν' 416 live, 
whence ζῆν and Bios. (Cf. also the probably Iranian proper name which 
in O. Pers. would be U-frdta, Εὐτφράτης, gs. *evrdnros well-filled.) 
Εὐαγγέλιον depends on the Descriptive evayyedos well-reporting, good- 
news-bringer. Though its application to the news itself instead of the 
messenger’s reward or the thanksgiving to the gods is not found in classical 


1(Mr. E. E. Genner observes that Welsh Owen is commonly derived 
from Eugenius direct, as Emrys=Ambrosius, where an “ Ancient British ”’ 
etymology seems impossible. —-Ep |] 


282 A GRAMMAR OF NEW TESTAMENT GREEK. [ὃ 106 


Greek, the more general sense of this derivative is current in ordinary 
Hellenistic and may be independent : it may even be a Descriptive taken 
directly from εὐ-αγγελία, with the decl. changed. [Εὐδοκία seems to 
start from εὐδοκέω, which will be treated under Cl. VII.] ἙΕὐωχέω 
(whence συνευωχεῖσθαι) may be classed here if it originates in a epd. 
εὔωχος (not found): this would be from «i -+-oxos (gradation of ἔχω), 
with the vowel lengthened in composition on analogy (cf. on χορηγύς, 
p. 275)—the meaning would be treating well. Other epds. of εὖ belonging 
to Cl. IV. are εὐάρεστος well-pleasing, εὐγενής well-born (or Cl. V. having 
good “ birth”’—see p. 287), εὔθετος well-placed, εὐλαβής handling well, 
cautious, reverent, εὐμετάδοτος readily distributing, εὐπάρεδρος assiduous, 
εὐπειθής obedient, εὔπορος well-faring (hence “péw and “ρία), εὐσεβής 
pious, εὐχάριστος grateful, εὔχρηστος serviceable. Εὐποιΐα need not 
imply a non-existent εὔποιος well-doing, but may be formed on analogy. 
Εὐτράπελος versatile, hence witty and (in mal. part.) ribald (whence “λία) 
comes from τρέπω : the adj. *rpdzedos is not found, but is guaranteed 
by its Latin equivalent torculus (trq“elos)—both=turning, but they are 
applied in different ways (Brugmann Gr.* 231). 


Εὐπερίστατος, dA. in Heb 12%, is a well-known crux. To the 
material given in Westcott’s note need be added only the fact that in 
vernacular Greek (see Deissmann BS 150) περίστασις often means distress, 
evil circumstances: we sometimes use circumstances with the same sugges- 
tion. This is what Theophylact was thinking of in explaining «v7. 
ἐς because of which one easily falls into distresses (περιστάσεις). I do 
not quite understand Westcott’s objection to Chrysostom’s “ what can 
easily suffer περίστασις i.e. removal.” “ΤΠ epd. could not lose the -ἰ- : 
it must be formed from orarés.” No doubt, but we should expect 
περίστατος and “ous to have kindred meanings. We are at liberty to give 
the verbal in -ros either active or passive force, the root being intransi- 
tive (Proleg. 221 f.). From the form it does not seem that there is any 
fatal objection to (1) easily avoided, (2) admired (lit. well-surrounded), 
(3) easily surrounding, besetting, or (4) dangerous (lit. having easy dis- 
tress). This last (Theophylact’s) implies that ro περίστατον got the 
sense of περίστασις in malam partem, so that the epd. becomes a Possessive 


(based on Cl. [V.a). 

Εὐθυδρόμος (whence “έω) may be put here as a cpd. of εὐθύ, straight- 
running. But as in other cases where an adj. seems used adverbially, 
we might get the same result by applying the Possessive, based on a 
Descriptive straight run. 

‘Hpbavys—=half dead: we can hardly class this with the numeral 
cpds. 

Μογίλαλος speaking with difficulty, and παλιγγενεσία re-generation, 
are both simple exx. of the present class. 

Ταλαίπωρος has in the first place a form from the root of τάλας ete. 


which seems to recur in the Gothic pulains, patience: ct. ταλαίφρων beside 
ταλάφρων, The πωρός presumed for the second half—apparently the 


§ 106-107] WORD-COMPOSITION—NOUNS. 283 


gradation of πηρός disabled—does not show itself outside the notes of 
ancient scholiasts and grammarians, who may be etymologising. 


Τηλαυγής far-shining or far-discerned, according as αὐγή or αὐγάζω is 
to guide our interpretation of the second part. 


δ 107. V. Possessive Compounds are 
called in Skt. Bahu-vrihi (B.V.), a term which 
illustrates the class: as a K.D. it would mean “ much rice,” 
but it has developed instead the possessive sense, “ having 
much rice.” These cpds. can be based on any of the foregoing 
classes, a distinction being made in the earliest times by change 
of accent. Thus in Skt. raja-putrad (Rajput)=king’s son, a T.P. 
(Cl. I.) ; rajd-putra=having a king for son, a B.Y. So in 
Greek Onpo-rpopos=feeding wild beasts (Cl. III.), while @npo- 
tpopos=having wild beasts as food (Cl. V.). Brugmann (Grd.* 
i. i. 75) shows that the accenting of the first element in a 
Possessive cpd. goes back to I.E. though comparatively few 
traces are left in Greek: this is a natural consequence of the 
Greek restriction of the accent to the last three syllables of 
a word, a restriction unknown to Skt. and to Germanic in its 
earliest stages. Brugmann, however, insists that the Possessive 
class is not to be regarded as secondary: it is as old as the 
other classes. 


Possessive. 


Based on Cl. I.—Possessives related to Cl. I. can hardly be illustrated 
from the NT, unless we treat νυχθήμερον as the adverbial acc. of an adj. 
=tncluding a day and a night. Here as in many other Possessives the 
line separating this class from others is rather unreal. 

On Cl, III.—’Aypavdos (whence “éw) is perhaps best taken as =with 
dwelling (αὐλή) in the fields: it might, however, be Cl. IV. 

Adroxetp is probably based on a Cl. III. noun hand of himself, hence 
with his own hand. 

Cpds. ending in -ὠδης (γραώδης anile, δαιμονιώδης devilish, θειώδης 
of brimstone, πετρώδης rocky) belong to this class if Wackernagel is right 
(see Giles Manual? 349) in comparing this suffix with that which meets 
us in εὐώδης (whence “ δία) fragrant, from ν' od of ὄζω, oleo, odor, Sew dys 
thus=having scent of brimstone, and the suffix lost its special limitation 
through its likeness to the -ῴδης produced by contraction of -o- with -ειδής 
(from εἶδος). 

On Cl. IV.—The great majority of Possessives in the NT are closely 
connected with the Descriptives, from which it is often impossible de- 

cisively to separate them. 


1 [See also p. 376 below, and Debrunner’s note, Wortb. 195 n.1.—Ep. | 
VOL. II. PART III.—20 


284 A GRAMMAR OF NEW TESTAMENT GREEK. — [§ 107 


To IV. (a) belong the following :—Atoypokepdys having base gain. 
As in the case of πλεονέκτης above, there is a desiderative tone imported, 
for the word is normally turpi-lucri-cupidus (as Plautus renders it). 

[Aicxpodoyia is best taken from αἰσχρολόγος, which may be analysed 
turpia loquens, Cl. IIL., or having foul speech, Cl. V. So patatoddyos 
tdle talker. ] 

᾿Αλλογενής of another γένος, and ἀλλόφυλος of another φυλή. 

᾿Αριστόβουλος (pr. name)=having excellent counsel. 

Βαρύτιμος --λαυΐηρ heavy price. 

ἘΒραδύπλοος may be assumed for βραδυπλοέω having a slow voyage. 
(Here as in other such cases we do not assume that the postulated word 
really existed: when an analogy was started, words would often skip a 
stage.) 

‘EtepdyAwooos=using a strange language. [The other two ecpds. of 
ἕτερος seem to belong to Cl. III., viz. ἑτεροδιδάσκαλος (whence “héw) 
teacher of strange things, and ἑτερόζυγος (whence “yéw) yoked to a stranger.) 

Ζώπυρον (whence “péw)! spark, or bellows, seems to be from (as, 
=what has live fire: hence the verb means kindle to flame. 

Ἱεροπρεπής having sacred seemliness, reverent, might be Cl. III. (dat. 
dependence) beseeming what 1s sacred. Μεγαλοπρεπής having great 
seemliness, majestic, belongs here. 

KaxonOns (whence “Se.a)=having evil nature, malicious. 

Μακροχρόνιος having long time, has an additional adjective suffix 
-cos—a superfluity not uncommon in these epds., both in Greek and 
Skt. 

Μετριοπαθής (whence “0éw) is the antithesis of ἀπαθής, Ξεμέτρια τὰ 
πάθη ἔχων. So ὁμοιοπαθής with like nature. 

Three cpds. of ὀλίγος come here, viz. ὀλιγόπιστος Little-faith, 
ὀλιγόψυχος little-souled, Faint-heart, and ὀλίγωρος (whence “ρέω, from 
ὥρα) little-caring. 

Ὅλος forms ὁλόκληρος having parts entire, and ὁλοτελής having entire 
completeness. Like the last is παντελής. 

From the base ὁμό- (whence ὁμοῦ, ὅμοιος etc.,=E. same—gradation 
form of sem- (eis) and sm (d-raé etc.)) come ὁμόθυμος (whence “μαδόν) 
one-minded ; ὁμόλογος (whence “γέω, °yia) having one speech, and so 
agreed (or same-speaking, Cl. III.) ; ὅμορος (whence cuvopopéw) having 
same boundary; ὁμότεχνος having same trade; ὃδμόφρων of one mind. 
With the last cf. σώφρων having sound mind and ὑψηλόφρων (whence 
“φρονέω (high-minded). We find in papyri ὁμότυπος, having one impression, 
i.e. duplicate copy. 

Ὀρθόπους (whence “ποδέω) having straight feet. 
Πάνοπλος (whence “ta) having full armour. 
[Παντοκράτωρ, if it means having all strength, belongs here. But 


1 [ἀναζωπυρέω 2 Tim 15. Simplex not in NT.—Ep.] 


§ 107] WORD-COMPOSITION—NOUNS. 285 


the termination is probably adjectival, a gradation variant of κρατερός, 
so that the cpd. goes into Cl. IV., All-mighty.] 

Πιθανολόγος (whence “ta) having persuasive speech, has the same 
doubtful classification as the other -λογος cpds. 

Πολλαπλασίων having many folds, manifold, has extra suffix -.os, and 
the individualising -ων : its base πλατ is from pjt, whence our fold. Other 
cpds. of πολύς have the base πολυ-. So πολυλόγος (whence “ia) much- 
speaking ; πολυμερής (adv. only) having many parts ; πσλυποίκιλος having 
much varied ; πολύσπλαγχνος having much compassion (so εὔσπλαγχνος) ; 
πολυτελής having great cost, so πολύτιμος ; πολύτροπος (adv. only) 
having many ways. 

Πραὔπαθής (whence Gera) having gentle nature : cf. the other epds. of 
πάθος. 

“Ῥᾳδιουργός (whence Ὕημα and ‘yia)=having easy (and so recklese) 
works: it might equally well be Cl. III. doing reckless things, or Cl. IV. 
reckless worker. 

Σκληροτράχηλος stiff-necked is like σκληροκαρδία (see above, p. 279). 

ExuOpwrds=with gloomy face. 

Σῶς, from σάξος (=tvavos, cf. Skt. tuvi “ strong’’), answers in mean- 
ing to sanus, “ sturdy.” It forms a good many proper names, as Σωσθένης 
from σθένος : Σώπατρος having a healthy father (or, having his father safe), 
will be a name given by the father because of auspicious omen for 
himself. Zaégdpwv—=healthy minded: cf. other cpds. of φρήν above, p. 284. 

Χρυσοδακτύλιος =having gold δακτύλιον (ring or rings). 

Ψευδής or its root makes in this class ψευδολόγος with false speech 
(but cf. above) and ψευδώνυμος with false name. 

To IV. (6) belong the following (as explained above) : 

One (1) sm, weak grade of sem (cis=sems, pia=smia etc.). 

᾿Αϑελφός, like ἄλοχος etc., has lost its breathing: there was a powerful 
analogic influence in d- privative, and probably also in a- from y, 
the reduced form of the prep. en. It is sm-g"elbhos “‘ having one 
womb.” 

["Axépatos has been plausibly explained as sm-kera-ios ‘‘ having one 
growth,” differing only in gradation from Lat. sin-cérus. But the ety- 
mology has been disputed by such authorities as Brugmann and Osthoff : 
it may be safer (with Boisacq Dict. Htym. 35) to interpret wndamaged 
(cf. κεραΐζω ete.), or again wnmixed (κεράννυμι). 

᾿Ακόλουθος (whence “€w) is sm-+KéAevOos path, so having one way, 
on the same way. 

Ἅπας, Skt. gacvant (=sa-gvant by assimilation), is a stronger form of 
πᾶς. 

᾿Απλοῦς (cf. διπλοῦς, Lat. duplus, Ger. Zweifel (‘double mind,” 
doubt)) has a somewhat doubtful second element (see Hirt Gram. § 372). 
Perhaps the element plo (seen in Latin) is increased by different suffixes, 
-vo- in Gk. -πλόβος, -to- or -tio- in Greek - πλάτιος (διπλάσιος, πολλαπλασίων) 
and Gothic ainfalbs (our -fold), -k- in Lat. simpler. A further element 


286 A GRAMMAR OF NEW TESTAMENT GREEK. [8107 


is seen in ἅπαξ, presumably a case (which, is not certain) of a cpd.= 
having one thrust (πήγνυμι fix): οἵ. Ger. einfach. 

(2) Μόνος gives us μονόφθαλμος with one eye. Μονογενής (see above) 
might also be classed as meaning of single birth. 

Two takes in Greek cpds. the form 6:- (=dFfi-, formed in prehistoric 
times on analogy of tri-; cf. Lat. bi-=dut-, etc.). So δίτδυμος twin (cf. 
ἀμφί-δυμος entered on both sides, νήδυμος enfolding (sleep): the root is 
that of dve—see Brugmann in /F' xi, 283,” and below, p. 305, on ἐνδύνω)--- 
δίδραχμος of two drachmae (see p. 280), διετής of two years (ib.), διθαλάσσιος 
(with added suffix) with two seas, δίλογος double-tongued, διπλοῦς double 
(see above), δίστομος two-edged (στόμα as in Heb 115 al.), δίψυχος with 
two selves. 

Three is τρι-, in τρίβολος (three-pointed, so as a noun) caltrop, τρίκλινος 
(whence ἀρχιτρ.) (a room) with three couches, τρίμηνος three months long, 
τρίστεγος with three storeys, or belonging to the third storey (a noun in 
Ac 20°). 

Four occurs in its most reduced form (4 "7, rrpa-) in τράπεζα four- 
footed (table). Less primitive epds. show rerpa- (q“etvy, cf. Gothic fidur-) : 
80 τετράγωνος fowr-cornered, τετράμηνος four months long, tetpat ots 
fourfold (see above), τετράπους four-footed. A curious analogy-product 
results from this last: the association of domestic animals and slaves in 
the household gave birth to ἀνδράποδα to stand by rerpdroda—in such 
‘a sense the word was first wanted in the collective plural: cf. the 3rd 
460]. dat. ἀνδραπόδεσσι in its earliest occurrence, /liad vii. 475. (Hence 
ἀνδραποδίζω ‘ to kidnap,” and noun ἀνδραποδιστής.) 

Eight from Hesiod down makes epds. in ὀκτα- (ef. revra-, €£a-, follow- 
ing ἑπτά etc.). So the NT ἅ.λ. ὀκτα-ήμερος (Ph 35) on the eighth day, eight 
days old. 

Ten appears in Δεκάπολις (sc. χώρα) having ten cities : ὃ cf. ‘Exraxwpia, 
a district named in a 1.8.0. papyrus (Archiv v. 38). 

Twelve makes δωδεκάφυλος having twelve tribes. 

Forty, like 2 and 100, forms a cpd. with Feros year, τεσσαρακονταετής 
forty years old. 

Hundred, besides ἑκατονταετής, makes ἑκατονταπλασίων hundred fold. 
The base follows τριάκοντα ete. 

To IV. (c) belong many epds. with a- privative, of which only a few 
need be named. 

ἼΛβυσσος (βυσσός bottom, cf. βυθός), sc. χώρα, the bottomless place. 


1 Brugmann (JF xi. 283) thinks πλόος voyage was used in a metaphor— 
like ὁ δεύτερος πλοῦς “ a second string to my bow ’’—and then extended when 
its meaning was forgotten. 

2? Brugmann thinks the meaning was obscured in prehistoric times, like 
that of ἁπλοῦς. May not the connexion he himself makes with νηδύς womb 
have helped the specialising of its meaning ? 

3. Cf. for this ellipsis τὴν δεκάπληγον (sc. τιμωρίαν) P Par 5749037, 


§ 107] WORD-COMPOSITION—NOUNS, 287 


᾿Αγενής having no “ birth’? (in the conventional sense), 80 ignoble : 
opposite of εὐγενής. 

[Ἀκριβής may be mentioned simply to say that its etymology is 
not known: for some guesses see Boisacq.] 

᾿Αληθής from ,/ of λήθω, λανθάνω, having no mistake, true. 

“Avous mindless (whence ἄνοια) : cf. εὔνους etc. 

᾿Ανωφελής profitless, from ὄφελος : for the lengthening of the ο, cf. 
Boisacq, 732. 

᾿Απρόσκοπος without mpookorn, is either active not causing to stumble, 
or passive not stumbling, void of offence. 

᾿Αργός, originally ἀξεργός, without work. 

[Ασελγής (whence ἀσέλγεια) is of unknown origin. The popular 
etymology connecting it with θέλγω is probably responsible for the 
spelling ἀθελγία in BGU iv. 1024" (iv/v a.D.). | 

ἔλστοχος (whence €w) having no aim. 

᾿Ασχήμων, from σχῆμα with its final -my replaced by the long grade 
-mon, =without due form, unseemly. 

"Atotos having no place, strange: in Hellenistic developed into 
wrong. 

᾿Αφελής (whence “hétys simplicity) seems rightly connected with 
φελλίς, φέλλια (pl.)—the form is imperfectly preserved—=stony ground : 
ἀφελής appears in Aristophanes as an epithet of a smooth plain. 

᾿Αχρεῖος (the derivative ἀχρειόω drops « in NT) is formed from χρή, 
a primitive noun=use, with suffix -vos: cf. Ion. ἀχρήϊος. 

One word needs adding that is formed with né, viz. νῆστις having no 
eating (*éoris from ,/ ed), hungry. 

Passing from the negative words, we come to miscellaneous items. 

Avo- forms δυσέντερος (which may be assumed as base of ‘pia, “ριον), 
having the ἔντερα wrong: the word is late, and was very likely remade. 


Also δύσφημος (see below). 
Εὖ forms a good many, of which we can name without comment 


εὐγενής well-born (see on ἀγενής), εὔθυμος good-tempered, cheerful, εὔκαιρος 
well-timed, εὔκοπος with easy labour, easy, εὔνους (whence εὔνοια, εὐνοέω) 
good-minded, friendly, εὔοδος with prosperous path (whence “δόω make 
prosperous), εὐπιρόσωπος (whence “πέω) with fair show, specious, εὔσημος 
having easy signification, intelligible, εὔσπλαγχνος kind-tempered, com- 
passionate, εὐσχήμων respectable (see on ἀσχήμων above), εὔτονος (relive), 
well-strung, vigorous, εὔφρων (whence εὐφραίνω and εὐφροσύνη) happy- 
minded, merry, εὔψυχος (whence °yéw) of good courage), εὐώδης (whence 
“Sia) well-smelling. ἘΕὔδιος (fem. used as noun) contains the weak 
grade of Ζεύς (diéys, the personification of the bright sky)=having fair 
sky. Εὔφημος by etymology means with good φήμη, with auspicious 
sound, as opposed to δύσφημος (whence “pia and “péw) inauspicious. The 
words were not without a wider use, and in 2 Co 68 presumably bear this : 
in Ph 45" a suggestion of the earlier association may well survive. Εὐώνυμος 


288 A GRAMMAR OF NEW TESTAMENT GREEK. [§ 107-108 


well-named is a conspicuous example of εὐφημισμός : the left hand, which 
in Greek augury was unlucky, was called ἀριστερά better or εὐώνυμος lucky 
in the hope that it might answer to its name. 


§ 108. VI. A name for the next class is not 
easy to find: we may call it Verbal, faute 
de mieux. Its characteristic is that the determining element 
governs the other as a verb does. This does not prevent its 
having the characteristics of a noun or adjective. Since in 
Composition we deal mostly with stems that will serve Noun or 
Verb equally, the line is not easy to draw ; and there is a whole 
class, the Determinatives with accus. dependence, which has 
been dealt with above, but must be noted as logically coming 
here. 


Verbal. 


The class is divided according as the verbal 
element stands first or second. A. Cpds. with 
verbal element in the second place include (1) 
a few in which a group of noun+governing verb has been 
contracted into a single word. Thus twypéw starts from ζωὸν 
ἀγρεῖν “to take alive”: the stem—casus indefimtus as Hirt 
describes it, since it will stand in any case relation or none— 
is substituted for the full word. As Hatzidakis suggests (Hunl. 
227),1 we have a trace of this mode of word-formation when a 
second element with verbal function is reduplicated for the 
perfect, without regard to its history : ἱπποτετρόφηκεν occurs 
in Lycurgus (iv/B.c.), though τροφέω does not exist, and if 
it did would govern an accus. This class of course is only 
treated here because we must register exceptions to the general 
rule that epd. verbs, if not formed by joining a “ proper” 
preposition to an existing verb, can only come into being 
through a cpd. noun or adjective. 


(A) Verb 
Second. 


It is likely enough that there are some other exceptions not provided 
for in our general theory. Thus such a word as ἀποκεφαλίζω, used in 
secular Hellenistic (Philodemus, Arrian), but not found earlier than in 
LXX, may be got from a presumed *dzoxépados having the head off. But 
the analogy of our own behead (by+head) suggests the probability of a 
direct coinage without this intermediary: the preposition then plays 


1 | Hatzidakis cites “ Lykurg. 139 nach allen Codicibus.” Mr. E. E. Genner, 
who points out that the latest Teubner emends it to the correct form, contests 
the attribution to Lyc. himself—Ep. | 


§ 108] WORD-COMPOSITION—NOUNS. 289 


the part of a verb (cf. ἄνα up /), and the cpd. gains resemblance to those 
in Class B below. (See, further, § 109.) 


(2) The bulk of the cpds. in this division are the Cl. III. 
words in which the second element has a strongly verbal foree— 
such as φωσεφόρος, Luci-fer, light-bringer. This noun is in fact 
an agent noun, or sometimes an action noun, as we have day- 
spring. It is largely a matter of indifference whether we do or 
do not transfer here a considerable number of the Cl. III. 
words, and not only those with accus. dependence. Cave- 
dwelling and cave-inhabiting cannot be classed apart, though 
cave in the first represents a locative and in the second an 
accusative. Brugmann notes (Grd.* τι. 1. 63) that agri-cola may 
be taken either as a Verbal cpd.=agrum colens or as a Deter- 
minative=agri cultor. It will not be necessary to deal further 
with this class, as the principle of it may be borne in mind 
while placing the Determinatives. 

B. Cpds. with the verbal element standing 

Oe first. There are two formations of this order 

ποῦ which come down from I.E. times, though 
they are not productive of new formations in Hellenistic. In 
(1) the verb stands in the same form as the imperative: thus 
"Apxé-haos ruling the people—the same sense can be obtained 
from a Cl. IIT. form, as πολιττάρχης. In (2) the form coincides 
with the stem of an action-noun in -ἰ, as μεμψί-μοιρος com- 
plaining of one’s lot. These action-nouns supply infinitives 
in Vedic Skt., and in Latin they can govern an accus. (cf. 
hanc tactio in Plautus). Brugmann (Gird.? τι. i. 64) regards 
them as infinitives with imperative function, so that both 
classes start from an exclamatory phrase. Such a noun as 
our skinflint could obviously arise from a sarcastic imperative 
turned into a nickname: cf. lie-a-bed, cut-throat, knock-kneed 
(with adj. suffix), twmble-down, forget-me-not, as varied exx. of 
the way such cpds. could originate. 

The form of words in the first division has been largely 
affected by the strength of the tendency to make o the vowel at 
the juncture. This goes with the fact that some pure noun- 
stems acquire im composition a strongly verbal meaning. 
Mioo-, vixo-, τῖμο-, φίχο- are not verbal bases at all: the 
verbs connected with them are secondary derivatives. Strictly 


290 A GRAMMAR OF NEW TESTAMENT GREEK. _ [§ 108 


speaking, pAddeApos—having a loved brother, Cl. IV. This 
classification will not suit the other three bases named : it seems 
best to suppose that since φιλάδελφος so clearly meant φιλῶν 
ἀδελφόν (or -ovs), analogy formed μισάδελφος as its opposite, 
and so the type grew. So Aristophanes coined Βδελυκλέων to 
match his Φιλοκλέων. We will include words of these types 
with the rest which agree with them syntactically. In some 
words the elision of the vowel makes it doubtful whether we 
have the old type or the new. 


(1) NT words of the first form are as follows: 


“Aypumvos (whence “véw “via) is supposed by Wackernagel to mean 
originally sleeping in the field (Cl. 111.}, like ἄγραυλος, ἄγροικος ; but 
the meaning sleepless is established from the first, and the connexion is 
strained. I prefer to take a base dypo- from the noun ἄγρα chase, and 
give it a verbal meaning, which may very well differ from that of the 
independent derivative verbs ἀγρεύω and ἀγρέω. Chasing sleep is 
sufficiently near to chasing sleep away. 

᾿Αρχέλαος has been already mentioned, the only NT word to preserve 
the primitive dpye- form (dpyéxaxos ete.). See above, p. 277, for this 
family. There are two words in which the first element is necessarily 
verbal, ἀρχισυνάγωγος and ἀρχιτρίκλινος : had these been formed in the 
earliest period, they would perhaps have had ἀρχε-. 

Βερνίκη is a Macedonian name also found in the form Βερενίκη, 
answering to Greek Φερενίκη carrying off the victory 

᾿Εθελοθρησκία, the well-known crux in Col 2%, may be paralleled 
with the conjectural ἐθελοταπεινοφροσύνη which Hort would restore 
in v."® for the almost unintelligible θέλων ἐν ταπεινοφροσύνῃ : Peake 
practically approves. If so, the ἐθελο- prefix tinges ταπεινοφροσύνη in 
v.*3 as well. Our would-be gives the sense adequately: Paul’s meaning 
will be rather like that of Jas 155, For other ἐθελο- epds. see LS: note 
the survival of the older form with initial ε, not found in Hellenistic 
with the verb θέλω. 

Μυωπάΐζω (2 Pet 1°, practically ἅ.λ.) is formed from μύωψ (Aristotle— 
μυωπός in Xenophon) =short-sighted, i.e. one who screws up (pie) his 
eyes tosee. For the formation see J. B. Mayor in loc. 

Νικάνωρ and Νικόδημος and Νικόλαος are names of almost the 
same signification, based on the verbal νῖκο-, conquering, on which see 
above. [Νικόπολις is better referred to Cl. IIL, city of victory.] 

Πείθαρχος (Aisch., whence “χέω) -- πειθόμενος ἀρχῇ, obedient to com- 
mand: cf πειθάνωρ, where likewise the noun is in dat. dependence. 

Τιμόθεος -ετιμῶν Θεόν : on the τῖμο- prefix see above. The name is 
classical, like its correlative Θεότιμος (Cl. III.), which might also mean 
God-honoured (instr. dep.). 


1 The second ¢ was lost by “* Kretschmer’s Law ’’—see p. 64. 


§ 108-109] WORD-COMPOSITION—NOUNS. 291 


The large φιλο- class has been also dealt with: it forms an immense 
number of cpds. in later Greek, some 15 of which occur in NT (reckoning 
only words with different second element). [Φιλόφρων (also adverb) is 
the only clear Cl. IV. word, =having friendly temper. | 

(2) From the second class we have but few : 

᾿Αλέξανδρος (pr. name) may be put here because of the antiquity of 
words like ἀλεξίκακος : either ἀλέκω or ἀλέξω (Skt. raks) will account for 
the form, which=defending men. 

᾿Ανεξίκακος forbearing evil, from ἀνέχομαι, is formed in the same way. 

Δεισιδαίμων shows stem assimilated as usual to sigmatic aorist : 
the root (ὠμοὶ to fear) has no -σις noun in Greek. Fearing the δαίμονες 
is the meaning: whether this comes nearer to religious or to superstitious 
is a question of usage. 

Δύω gives us the only others (μεμψίμοιρος having been dealt with 
above). Λυσανίας (pr. name) from dvia=destroying trouble. Λυσιτελής 
(whence “Aéw) paying expenses, and so profitable. The v is long, so that 
we have another ex. of the assimilation to the aorist stem: in λύσις of 
course it is short. That new cpds. could be made in this class may be 
illustrated from the Alexandrian tribe Σωσικόσμιος (as in P Strass 52?- 
151 a.p.),! called after the Emperor (Hadrian ?) as ἔσωσίκοσμος = σώζων 
τὸν κόσμον. Schubart (Archiv v. 99n.) mentions also two other tribes 
in Alexandria (both i/a.D.), Φυλαξιθαλάσσειος and Αὐξιμητόρειος, and 
(p. 101) Αὐξισπόρειος. 


§ 109. VII. Last we may define a class of 


Ῥ = Sead prepositional epds. which have apparently been 
Dies developed directly from a phrase: the exx. 


will be treated under the head of their respective 

prepositions in the next section. It is clear that we cannot 
force into previous classes such a word as πρόσκαιρος temporary, 
which is simply πρὸς καιρόν run into a single word and declined 
afresh. Often we have more complex suffixes added, as καθη- 
μερινός daily, from καθ᾽ ἡμέραν, ἐνύπνιον---τὸ ἐν ὕπνῳ dream. 
Similarly we find verbs made in this way, as κατακρημνίζειν-ε- 
to cast κατὰ κρημνοῦ--- ἀποκεφαλίζειν--ἀπο(κόπτειν)τὴν κεφαλήν. 
Closely akin to these are two formations in which the first 
part is an adjective, but the second part is dependent on it, 
instead of agreeing with it. Μεσοποταμία (properly an adj.) is 
ἡ ἐν μέσῳ τῶν ποταμῶν χώρα: cf mediterraneus, and our 
midland, if it means “ what is in the midst of the land ”—of 
course it might be “land which is in the midst.” The ἴσος 
cpds. have a dative dependence: ἰσάγγελος---ἴσος ayyéhous— 


1 (Cf. P Oxy iii. 5134 (184 a.p.).—Ep. | 


292. A GRAMMAR OF NEW TESTAMENT GREEK. [§ 109-110 


the only other two in the NT (ἰσότιμος and ἰσόψυχος) belong 
to Cl. IV. 


Like iodyyedos are ἰσοπλάτων (Anthology) equal to Plato, ἰσόνειρος 
(Aeschylus) like a dream, ἰσομάτωρ (Theocritus) like one’s mother, ἰσολύμπιος 
(Philo) like the Olympians, ἰσοκιννάμωμος (Pliny) like cinnamon, ἰσόθεος 
(Homer down) godlike, ἰσοβασιλεύς (Plutarch) equal to a king, ἰσάδελφος 
(Euripides) like a brother, ete. The formation was still capable of being 
made afresh in Hellenistic, and if Lk 20% gives us a new coinage, it is so 
entirely according to analogy and the practice of other writers that we 
cannot treat it as the basis of any inference as to Luke’s Greek. 


Εὐδοκέω (and -ta) stands apart as a new verb made with 
an adverb. There is no reason why we should postulate an 
adj. evdoxos: the verb has probably come straight from εὖ 
δοκεῖ ‘it pleases me well,” fused into a closer union by usage. 
Laws of speech were made for men, and not men for the laws. 


PREPOSITIONS. 


§ 110. The part played by Prepositions in Word-composi- 
tion is of such peculiar importance that it will be well to make 
a separate section of it, gathering here all that must be said of 
their formation and history, and their functions in composition. 
Their use with nouns must be reserved for the Syntax under the 
Cases ; and we shall also have to return under Verb-syntax to 
some of the phenomena of perfectivising (Prol. 111 f.). It will 
be convenient to anticipate the Syntax by indicating the cases 
with which the prepositions are found in the NT: the Gen. will 
have *, the Dat. 7, the Accus. f. 

It is usually assumed that Prepositions are simply Adverbs, 
separated from their class by special uses. Many of them are 
obviously specialised cases of nouns: some of the oldest, how- 
ever, have no link with existing cases, unless it be the casus 
indefinitivus which is used in noun-composition. Delbriick’s 
most important discussion, in his chapter on Prepositions in 
Grd. 11. (i.), makes it clear that in their origin they were 
not Adverbs at all in the ordinary sense. Their oldest use was 
as Praeverbia, and their function as prepositions “‘ governing ”’ 
cases was derived from this, as was also their rare use as simple 
adverbs. If this is so—and the facts of Vedic Sanskrit and 
Homeric Greek unite with scattered indications from less primi- 


δ 110-111] WORD-COMPOSITION—PREPOSITIONS. 293 


tive idioms to make it extremely probable—the old gram- 
marians followed a right instinct when they classed Prepositions 
as “‘ Proper’ and “ Improper.” The latter are ethnic develop- 
ments, not used in verb compounds, but only in the later 
function in association with nouns. There were even praeverbia 
which never became “ prepositions ” at all. These, as retaining 
only what we regard as the most primitive force, we may take 
as our first class, dealing next with the Prepositions, and finally 
with the “ Improper ”’ Prepositions, as latest in development. 

In distinguishing below (a) Composita proper and (6) com- 
pound nouns—often only seen in verbs derived from them—we 
note Delbriick’s argument that noun compounds began as mere 
adjuncts to Composita formed from the same verbal root. Even 
as late as Hellenistic this class is far more numerous than that 
which contains real nouns. Compounds belonging to Class VII. 
are apparently the earliest to arise. With these I have put the 
verbs of this class, as described above (p. 291). Miscellaneous 
derivatives go together in a class (c), where especially are found 
the compound prepositions (like our be-neath, Lat. s-wper= 
(e)x-uper, ἔναντι etc.) : In many of these we cannot strictly use 
the term compound, as the word is only a stereotyped phrase, 
starting from separate words in syntactical relation. These 
are as old as Homer, e.g. μετέπειτα, and are greatly extended in 
Hellenistic, 


I. PRAEVERBIA. 


§ 111. Hardly any of this class leave any 
traces in Greek, except those (forming our 
next section) which developed the use with cases, whether or 
no this use was in existence in the proethnic period. Thus 
we have no trace of Skt. sam “together”? (Av. ham, also in 
Lith.-Slav.), nz “down” (Av., and our nether, beneath: also 
cf. ne-st and Lat. nidus=ni-zd-os), vi “ apart” (Av., and Ger. 
wider), ava “oft” (Av., Lat. au-, and Slav.), ud “out” 
(v8pus—ef. βρι-αρός, ὕστερος, utter), ati “ beyond ᾿᾿ (also prep. 
in Skt., and adverb in ἔτει, Lat. οἰ). Only one of these is at all 
conspicuous in Greek, viz. w-, Skt. @ (also preposition).! It is 
well seen in ὠκεανός Skt. dgayana “ surrounding,” which seems 


Praeverbia. 


1 Brugmann Dem. 142 sees it in Lat. id-6-neus, Goth. it-a, also pan-a ete, 
(τόν - ὦ). 


294 A GRAMMAR OF NEW TESTAMENT GRERK. [§ 111-112 


thus to be a participle of ξὥκειμαι “lie around.” Ὠρύομαι (cf. 
Lat. rimor etc.), ὠφελέω (? Skt. phala “ fruit, profit’), ὀδύρομαι 
(whence ὀδυρμός) and ὀκέλλω (ἐποκέλλω late MSS in Ac 2741), 
compared with their simplicia δύρομαι and κέλλω, οἴγω (whence 
ἀνοίγω) from ὀξείγω, ὀδύνη (perh. from δύη 2), ὀμείρομαι (2,/ smer 
“to remember,” ‘‘ bethink ’’—Lat. memor etc.), ὄνειδος (1, nerd, 
newt, cf. Ger. Ned, O.K. nip). The shortened o- may point 
to gradation : cf. é- against ἡ- (below). The general sense of 
the prefix seems to be directive. But the whole of this entry 
is decidedly charged with uncertainty. The compounds with 
other prepositions will be seen in III. below. <A _ cognate 
preverb 7 (cf. Lat. δ) may account for ἤ-θελον. (See p. 188.) 


Il. PREPOSITIONS. 


§ 112. The Proper Prepositions are here 
discussed in their function as praeverbia, 
and in the probably derived use as forming 
noun compounds. The latter may be assumed to be of the 
IV.c class (p. 280 above) unless otherwise stated. Under each 
preposition (a) contains preverbial uses, (b) noun compounds, 
and both nouns and verbs of Cl. VII. (p. 291), while (c) 
contains adverbial compounds or phrases that have become 
stereotyped as single words. 

ἈΠ 1. ᾿αμφί on both sides, obsolete in Hellenistic 

es as a preposition.' Latin ambi-, OHG umbi 

(Ger. um), Skt. abhitas “ on both sides,”’ make its meaning clear, 

as does its connexion with ἄμφω ambo. Brugmann (KVG 468) 

regards it as a compound, the second part of which is compared 
with Eng. by (Ger. bei), and Skt. abhi, Av. ait. 


Proper 
Prepositions, 


(a) ᾿Αμφιβάλλω in NT is the verb of ἀμφίβληστρον, but in its oldest 
use was especially applied to putting on clothes : the dual idea suggested 
to us by the symmetrical shape of our garments applied less obviously 
in ancient times. The only other verb compound in NT is that for 
clothing, in its literary form ἀμφιέννυμι (Mt 68°) and vernacular 
ἀμφιάζω (Lk 1235). 


(6) The original dual meaning survives in the place-name ᾿Αμφίπολις 
on-both-sides-(the-river)-town. In ἀμφίβληστρον the meaning casting-net 


1 It is a dual word, and disappears before περί for reasons explained in 
Prol. 57, 77-80. 


§ 112-113] _ WORD-COMPOSITION—PREPOSITIONS. 295 


is as old as Hesiod, and the separate manipulation of the net’s two ends 


makes the ἀμφί appropriate. But ἄμφοδον (a road round) is like most 
of the compounds not much concerned with duality. 


atte § 113. 2. PAva upwards, as an independent 

word accented ava (notin NT). It is found in 

Avestan ana, Gothic ana, Slav. na, where the original wp is only 

partially visible : Lat. an-helare shows it well. Over, of space 

covered, on (as in the English cognate), and wp to, of a goal 

attained, are developments reached in other languages than 
Greek. See Delbriick Grd. m1. (1.) 734. 


(a) 71 composita are found in NT. Of these 13 show the local force 
up only: ἀναβαίνω go up, ἀναξώννυμι gird up, dvaxpdlw, ἀναφωνέω 
and ἀναστενάζω cry (groan) up (i.e. the sound is fetched wp by a deep 
respiration), ἀναπληρόω fill up, ἀνέχομαι hold oneself up, endure etc. 
In 26 ἀνά answers to Lat. re(d) in its rather different senses again and 
back. ᾿Αναζάω live again, revive, ἀναγεννάω regenerate, ἀνακαινίζω and 
-όω and ἀνανεόω renew, ἀνανήφω become sober again, ἀνοικοδομέω rebuild 
show the former: ἀνασταυρόω crucify afresh, which in classical Greek 
shows ava=up, illustrates the close connexion of all these meanings. 
For back cf. ἀναχωρέω retire, ἀνίημι relax, ἀνακυλίω roll back etc. A few of 
these answer to the English prefix wn- (Goth. and-, ἀντί), which reverses 
an action: dvakahumtTw=re-velo, un-cover, ἀνακύπτω (gs. wn-stoop) lift 
oneself wp, ἀνασκευάξω (qs. disfurnish*) unsettle. In some verbs these 
various meanings of dvd exist side by side: thus ἀναπέμπω send wp or 
send back, ἀνίστημι raise up or (make stand again) restore to life, ἀναβλέπω 
look wp or regain sight. Naturally there are ambiguous cases: thus 
Delbriick (Grd. m1. (i.) 738) makes évéxw—=hold back, restrain. 

A difficulty is raised by ἀναγινώσκω, which in the Κοινή has always 
the peculiar Attic meaning read (generally read aloud). In Homer 
avayvavar=recognise, and once or twice in Herodotus. But in the latter 
there is a transitive meaning persuade, found in pres. and Ist aorist. 
It looks as though a transitive force, cause to understand, came from the 
compound and developed independently in Attic and in Ionic: it may 
even be proethnic (not of course in the developed sense), since anakunnan 
in Gothic means read.2 There are other verbs in which ava either gives 
or preserves a causative force. ᾿Αναμιμνήσκω remind and ἀναφαίνω 
(nautical term) sight are transitive in the simplex; but not ἀναθάλλω 


1{Mr. E. E. Genner questions the meaning “disfurnish,” and observes 
that dvack. τὴν τράπεζαν in the Orators is just like our ‘“ reconstruct” a 
company (euphemism for bankruptcy).—EpD. ] 

3 But only in 2 Co 11* 32, where there is paronomasia with other cpds. of 


γινώσκω. Elsewhere ἀναγινώσκω is ussiggua. This rather weakens the in- 
ference. 


296 A GRAMMAR OF NEW TESTAMENT GREEK. [§ 113-114 


make to bloom, revive, nor (practically) ἀνατέλλω make to rise. ‘Che anti- 
thesis of ἀνά has this causative force among its functions (see κατά below), 
and it seems possible that ava may have developed it. 

In a considerable number of ava compounds we may recognise a 
distinctly perfective force. As was shown (Prol. 112) by evidence from 
English, perfective force may coexist with the local force of the preposi- 
tion, there being always a more or less strong tendency to make com- 
pounds perfective. “Ava is not one of the more markedly perfective 
prefixes, but such compounds as eat up, grow up, fill up, upset, done up 
(pte.) illustrate how naturally it may take this function. The following 
NT verbs may be classed under this heading :- -ἀναδείκνυμι (show up), 
ἀναδέχομαι (cf. taking wp an acquaintance); ἀναδίδωμι (hand in docu- 
ments), ἀνατίθεμαι (sim.); dvalntéw (cf. our hunt up references), 
ἀναθεωρέω. dvakpivw, ἀνετάζω, ἀνευρίσκω ; ἀναζωπυρέω (make burn up), 
ἀνάπτω ; ἀναιρέω (take up, remove, destroy), ἀναλύω (break wp), ἀναλίσκω 
(probably =dva-FaXioxw, so take wp, spend); ἀνακεφαλαιόω (sum up) ; 
ἀναμένω (cf. stay up for) ; ἀναπείθω (seduce—the perfective suggests success : 
ef. Prol. l.c.), ἀνασείω (shake up); ἀνοίγω (our open is akin to wp); 
ἀνατρέφω (bring up); ἀνήκω (come up, arrive, hence belong, befit). 

(b) Many of the nouns simply attach themselves to the verbs described 
above. For the local meaning add ἀνάγαιον wpper room, perhaps formed 
by antithesis from κατάγαιον ! cellar; the doublets ἀνάθεμα and ἀνάθημα, 
votive offerings hung up on a temple wall; ἀνάπειρος (cl. ἀνάπηρος), 1.6. 
maimed all the way up ; ἀνάστατος (whence the vernacular verb ἀναστατόω 
upset), with the suggestion of ἄνω κάτω upside-down in it. 

*Avadoyia is of a different formation, coming from the phrase ava 
λόγον proportionately : it belongs to Cl. VII. 

(c) On ἄνω see ὃ 130 s.v. ἔξω: hence ἄνωθεν, ἀνώτερος, ἀνωτερικός 
and ὑπεράνω (an impr. prep.). 


§ 114. ὃ. *Avri an front, opposite, is found 
in most of the LE. dialects: Skt. anti “ over 
against, near (adv.),”’ Lat. ante “ before,” Goth. and “along, 
on,” Lith. ant “on, to.” Cf. also ἄντα, Goth. anda- (as 
anda-bauhts=avritutpov). The local meaning which (with the 
consequent temporal) occupied the whole field in Latin appears 
in some Greek dialects.2_ Thus Cretan ἀντὶ μαιτύρων “before 
witnesses.”’ It survives freely in composition. We may note 
the verb avtdw (whence ἀπαντάω etc.) as formed from ἄντα : cf. 
our verb to face, also there is the adj. ἀντίος (Homeric), whence 
ἐναντίος, ὑπεναντίος. 


2 , 
Avtt 


1 Class. ἀνώγαιον is not quite clear (see pp. 70, 76). 
? And even (very rarely) in Hellenistic. 


δ. 114-115] WORD-COMPOSITION—PREPOSITIONS. 297 


(4) Opposite is the obvious starting-point in all the 20 composita of 
the NT. ᾿Αντιπαρέρχομαι pass by on the other side shows it in the simplest 
form. ᾿Ανθίσταμαι (Lat. antisto “‘ excel,’ Goth. andstandan ‘ oppose ’’) 
develops “‘ oppositeness” into “opposition”: so ἀνταγωνίζομαι, 
ἀντιλέγω (ἀντεῖπον), ἀντιδιατίθεμαι, ἀντικαθίστημι, ἀντίκειμαι, ἀντιπί- 
πτω, ἀντιστρατεύομαι, ἀντιτάσσομαι. Reciprocal action appears in ἀντα- 
ναπληρόω, ἀνταποδίδωμι, ἀνταποκρίνομαι, ἀντιβάλλω, ἀντικαλέω, 
ἀντιλοιδορέω, ἀντιμετρέομαι : in classical and Κοινή Greek the same 
sense attaches to ἀνθομολογέομαι, but in its NT occurrence, as in LXX, 
this is modified by the context—thanksgiving in return for benefits. 
᾿Αντέχομαι and ἀντιλαμβάνομαι go together: perhaps the local force 
survives, to express grasping while squarely facing the object. 

(b) Putting aside, as before, the nouns that answer to verbs given above, 
there are not many left. The original local force appears in ἀντόφθαλμος 
(only Hesych.) which may perhaps be presumed for ἀντοφθαλμέω : 
it might, however, be a Cl. VII. verb from ἀντ᾽ ὀφθαλμῶν. Opposition 
appears in ἀναντίρητος, ἀντίδικος, ἀντίθεσις ; reciprocity in ἀντάλλαγμα, 
ἀντίλυτρον, ἀντιμισθία, where is implied the equivalence of the object 
to that against which it is set. Equivalence in a different way is implied 
in ἀντίτυπος, ἀνθύπατος (=pro consule), ἀντίχριστος, which last is not 
“an opponent of Christ”? but ““one who assumes the guise of Christ ”’ 
in order to seduce His people, just as ἀνθύπατος is “‘ one who holds the 
power of a consul.” The proper name ᾿Αντίπατρος, ᾿Αντίπας should 
probably be classed here: cf.”Avravdpos in BGU iv. 11348 (B.c. 10). 

(c) The only adverbial derivatives of ἀντί appear in the list of Im- 
proper Preps. 


δ 115. 4. * And, off, away: as an independ- 
ent word accented ἄπο (not in NT). Skt. dpa, 
Av. pa, Lat. ab (cf. aperio) and po (in po-situs al.), Goth. af ete. : 
cf. also the adverbs ay abs (for the -s see pp. 164, 329). The 
meaning is practically the same all over the field, all the uses 
of the word starting from the notion of separation. 


᾿Από 


(a) 90 composita occur in NT, in about 40 of which the local force 
described above is clearly visible (with perfective force in some). Some 
of them are proethnic: thus ἀποτίθημι, ἄπειμι (abeo), ἀποβαίνω, 
ἀφίστημι, ἀπάγω may be recognised in at least two other. I.E. languages. 
The common meaning off or away produces different nuances according 
to the meaning of the verb root. Sometimes the starting-point is the 
subject, sometimes the object: ἀπαγγέλλω bring news (from) starts at one 


1¥or this use is not unknown in Hellenistic, where the verb arose: see 
ἀντί in the Syntax. But the other is perhaps more likely, as in compounds 
this local sense is obviously active. 


298 A GRAMMAR OF NEW TESTAMENT GREEK. [8116 


end, ἀποδέχομαι welcome and ἀπεκδέχομαι wait for (see ἐκ) at the other. 
*Atrettrov=renounce (cf. Goth. afaikan and afgipan, both translating 
ἀπαρνεῖσθαι). In ἀποβλέπω and ἀφοράω looking away to an object 
suggests concentration: we might as well class them as perfective. 
᾿Αποθησαυρίξζω store away has likewise practically perfective force, 
and so ἀπολούομαι and ἀπονίπτομαι wash away, ἀπομάσσομαι wipe off, 
ἀφίημι and ἀπολύω release, dismiss, ἀποφεύγω flee away, escape (φεύγω -- 
flee), ἀφορίζω and ἀποδιορίζω separate off, ἀφυπνόω fall off to sleep. ᾿Από 
is, as we saw in Prol. 112 ff., one of the most conspicuous of perfectivising 
prefixes: quite one-third of the NT composita have perfective force 
more or less clearly recognisable. It will be well to continue the list, 
premising that the similar use of the cognate off enables us generally to 
translate literally :---άἀσπάγχομαι and ἀποπνίγω (cf. choke off, and ἔπνιγεν 
simplex in Mt 1838), ἀπαλλάσσω (starting from the idea of a complete 
change!), ἀπαλλοτριόω (cf. abalienare), ἀπαντάω and ἀφικνέομαι,3 
ἀπαρνέομαι (cf. abnego), ἀπεκδύομαι, ἀπέχω (as used in receipts ὃ), 
ἀπογράφομαι (ἢ write oneself or one’s family off), ἀποδείκνυμι (cf. show 
off), ἀποθλίβω (unless comparable with ἀποδέχομαι above), ἀποθνήσκω 
and ἀπόλλυμι and ἀποκτείνω, ἀποκλείω (shut off), ἀποκυέω (contr. κυέω -- 
be pregnant), ἀπορφανίξζω and ἀποστερέω, ἀποστυγέω, ἀποτελέω, ἀποτολ- 
paw (carry daring to its limit), ἀφομοιόω (finish off the likeness). In 
ἀφυστερέω the perfectivising preposition apparently produces transitive 
force. There remain two other developments of the local ἀπό. In several 
verbs we render back. ᾿Απαιτέω is to demand back one’s own, ἀποδίδωμι 
to give back what belongs to another, ἀπολαμβάνω to receive back, ἄπο- 
τίνω to pay back, ἀπονέμω to dispense back, assign, so also ἀποκαθιστάνω 
to establish back, restore, ἀποκρίνομαι to decide for oneself back, reply, 
ἀποκαταλλάσσω to effect a thorough change (perfective xara) back, 
reconcile. In another class the ἀπό reverses the verb’s action :—amahyéw 
have pain off, be past feeling, ἀπελπίξζω cease hoping, ἀποδοκιμάζω dis- 
approve, ἀποκαλύπτω take off covering, unveil, ἀποστεγάξω wnroof, amo- 
φορτίζομαι unburden, ἀποψύχω cease breathing, faint. 

(Ὁ) Some twenty of the noun compounds of ἀπό go with verb com- 
pounds accounted for under (a). Among these ἄφιξις departure alone 
need be specially mentioned, as having a peculiar meaning, divergent 
from that of its verb ἀφικνοῦμαι arrive: see Prol. 26 n. The noun 
ἀπαύγασμα (Wis 725, Heb 1%) comes from ἀπαυγάζω, where we have to 
choose between dzré=from, away and aré=back, between radiance and 
reflexion, sunlight and moonlight. Philo’s usage is divided. The Greek 
Fathers are unanimous for radiance: see Westcott’s citations in loc. 
The Biblical use—effulgence, pas ἐκ dards—seems to me fairly certain, 


1 Note the very common Hellenistic use of the pf. pte. ἀπηλλαχώςΞεαεαά. 

2 Cf. Prol. 247 on the late development by which ἀπέρχομαι was similarly 
transferred to the goal. 

3 Prol. 247. 


§ 115] WORD-COMPOSITION—PREPOSITIONS. 299 


though the RV should have given a marginal alternative, as the question 
is exegetical rather than grammatical or lexical, and on the latter grounds 
the choice is quite open. Other words showing local ἀπό are ἀποστασία 
(cf. ἀφίσταμαι), ἀποστάσιον (cf. ἀφίστημι and see Vocabulary s.v.), ἀποτομία 
(from ddéropos=abruptus), ἀφορμή (what one starts from). A partitive 
sense is given by the prefix in ἀπαρχή (something to begin with taken 
from the whole). ᾿Απελεύθερος freedman (édedPepos=freeman) con- 
notes reparation from a former master: ἀπολύτρωσις redemption is 
parallel. ᾿Από is back in ἀπολογέομαι (like ἀποκρίνομαι), but there is a 
difficulty about the formation, as ἀπόλογος, from which it would seem to 
come, has not the meaning defence (an argument back). Our classing the 
word in (a) is forbidden not by the fact that Noyéw does not figure in our 
dictionaries, which might be accidental: it does not seem that the word, 
even if it existed (which is very unlikely), gave birth to ἀπολογεῖσθαι as 
a genuine compound. Either (1) we must assume that ἀπόλογος once 
existed, probably as an adjective, and went out of use in this sense after 
producing ἀπολογία and dmodoyéouat—a fairly easy supposition. Or 
(2) we might suppose the verb formed directly from ἀπό and λόγος, 
just as ἀποδιδόναι λόγον might have been the correlative to αἰτεῖν λόγον 
(1 Pet 315): this would bring the word in line with ἀποκεφαλίζω and 
others below. It should be added that LS would take ἀπό as away, in 
the sense of repelling an accusation from oneself. ᾿Απαρτισμός com- 
pletion comes from the perfective verb ἀπαρτίζω, and ἀπόχρησις from 
ἀποχράομαι use wp. ᾿Απελεγμός disrepute is like ἀποδοκιμάζω; in 
ἀπόλαυσις from ἀπολαύω enjoy the prep. points to the source from which 
the enjoyment comes. Similarly ἀποκαραδοκία is parallel with ἀποδέχομαι 
(above): for the second element see above, p. 274. 


(c) There are a few compounds that derive from a phrase (Cl. VII.). 
᾿Αποκεφαλίζειν was explained above (p. 288). ᾿Απόδημος (whence 
ἀποδημέω) is simply ἀπὸ δήμου. The formation was still capable of use 
for new words, as amoguvdywyos—much as we could coin verbs like 
un-church as the need arises. “Adedpos, ἃ Kown word found in LXX 
(whence ἀφεδρών) is from ἀφ᾽ ἕδρας (LS s.v. ἔδρα iii.). A verb formed 
like ἀποκεφαλίζω is ἀποστοματίζω, which is in earlier Greek exactly like 
eatemporise, even to the verbal suffix (extempore applied to speech is ἀπὸ 


στόματος). In NT the meaning is factitive—to force to extemporise, 
to heckle. 


᾿Απέναντι appears among the Improper Prepositions. “Amdptt, or 
if preferred ἀπ᾽ ἄρτι (WH)—the Greeks would hardly have recognised the 
difference—is no true compound, for ἀπό practically governs ἄρτι as a 
case (cf. ἀπὸ τότε). It is only named here to distinguish it from the 
classical ἀπαρτί exactly (glossed ἀπηρτισμένως, τελείως, ἀκριβῶς), ΟΥ (as 
some said) on the contrary, which was a real compound, with different 


1 Thayer-Grimm is quite wrong in postulating στοματίζω, which neither 
does nor can exist. 
VOL. 11. PART III.—21 


300 A GRAMMAR OF NEW TESTAMENT GREEK. [§ 115-116 


accent. LS recognise the Hellenistic meaning in Aristophanes and 
Plato (Comicus), which is of course just the one kind of Attic wherein a 
Kown formation might be foreshadowed.t But see the discussion in 
Lobeck Phryn. 20 f. 


§ 116. 5. *tAvé through, orig. between through, 
is apparently cognate with Lat. dis- (Ger. zer-). 
This equation is simple and satisfactory, the only divergence 
being that Greek has added -a (ἔδισ-ἀ): Brugmann KVG 478 
thinks the analogy of μετά has been working. The question is 
whether we may identify this ἔδισ(ά) with the δίς or δίς which 
comes from I.E. dui, our twice (cf. between for connexion of 
meaning). Greek would of course leave this quite open. But 
I.E. dyis is bis in Latin and *twis- in Gothic (Ger. zwischen). It 
is proposed to postulate I.E. doublets dyis and dis, postcon- 
sonantal w disappearing under presumable sandhi conditions. 
See the evidence in Brugmann Grd.? τ. 259, where the proof of a 
sound-change in proethnic I.E. is regarded as indecisive. There 
seems to me a balance of probability in favour of our regarding 
this dropping of wu as proethnic rather than independent : per- 
haps in either case we may accept it for the several dialects, and 
so connect διά with the numeral two. Hirt Gram. 219 notes 
on διά, “ es steht wohl fiir δ(ε)ισα." Outside Greek this adverb 
does not become a preposition. 

Διά forms 79 composita in NT: it is noteworthy that 200 
out of 343 occurrences of these are in the Lucan Books. The 
διά compounds offer rather special difficulties, from the fineness 
of the distinctions between the classes in which we may place 
them.? The survey of the whole field shows us that the ety- 
mological connexion with two justifies itself by usage. To 
represent it graphically, we have two poimts or areas (A) (B) 
set over against one another, and the preposition is concerned 
with their relations and the interval between them. Per, 
trans and inter, as well as dis, will describe different relations, 
while our own through and through (per) comes with equal 
naturalness : per covers the intervening space in passing from 


1 [The new LS remarks after the Hellen. citations, “ This is not an Att. 
use, hence Pl. Com. 1. 43 must be incorrectly interpr. by” Anecdota Graeca 
(ed. Bekker.) 79.—Ep. ] 

2 Winer’s monograph (part v. of his De Verborum Compositorum in N.T. usu, 
1848) is still most valuable, though of course needing supplement now. 


Διά 


§ 116] WORD-COMPOSITION—PREPOSITIONS. 301 


A to B, trans only considers the reaching of B from A, while 
inter stops on the road, and dis still more emphatically dwells 
on the interval as a gulf fixed between them. Dis accordingly 
coincides only with part of dva’s area, and does not so often 
include the perfectivising force which is conspicuous in διά 
compounds. This is well brought out by the opposite meaning 
of διαξζώννυμαι (perfective) gird right round. and  discingor 
ungird: the sense of reversal can attach to διώ (not in NT), 
as the equation diafevyrvupar=disiungor will show. In the 
attempted classification below we have alternative possibilities 
presented to us in many cases. We can in some of these only 
come to a tentative choice, depending generally upon the 
relative antiquity of meanings attaching to the word in question, 
where a meaning appears to be derived from earlier develop- 
ments of the word itself, rather than from independent applica- 
tion of varying uses of διά to the word with which it is 
compounded. 


(a) We may take first composita in which dua=per. These are almost 
necessarily perfective, though the original force of the prefix is retained, 
for they describe the carrying of action through to a definite result. 
Spatial διά may be seen in διαβαίνω et sim. (five others), describing 
some sort of a journey to a goal. δΔιηγέομαι is the same used meta- 
phorically ; διασῴζω is to bring safely through, and διαφεύγω, διαφυλάσσω, 
διατηρέω are similar. Διανύω and διατελέω, διαπρίω (metaph.) and 
διορύσσω and διυλίζω, διαγγέλλω and διαφημίζω and διαγογγύζω, δια- 
καθαίρω and διακαθαρίζω all show space traversed; διακούω, διαμένω, 
διατρίβω, διανυκτερεύω, with temporal διά, may be reinforced with 
διαγρηγορέω if it means remain awake throughout (but see below). 
Διαυγάζω (ultimately from αὐγή, the Hellenistic for ἕως) has probably 
local διά, of dawn breaking through: here Latin has dis (dilucesco). But 
it is obviously not far from διαβλέπω, which we might class as a pure 
perfective. Temporal διά more in the sense of inter is seen in διαγίνομαι 
intervene (of time) and διαλείπω (-λιμπάνω in Ac 85: D). 

Trans will perhaps be the closest equivalent for διά in διαβάλλω = 
traduco, durchziehen, as it were to toss across: Winer compares διασύρω 
(pull to pieces) which would suggest dis also for διαβάλλω, gs. “ throw to 
the winds.” Winer compares διαχλευάζω verspotten, which I would 
rather put below. δΔιερμηνεύω is translate. Through has passed into 
thoroughly in the pure perfectives διαπραγματεύομαι (Prol. 118), διαφθείρω, 
διασείω (? give a thorough shaking to, hence extort by intimidation—Lat. 
concutio), διεγείρω, with which we might class διαγρηγορέω if taken as in 
RV teat Lk 932 (with ingressive aorist). Διακατελέγχομαι (ά.λ. in Ac 188) 


302 A GRAMMAR OF NEW: TESTAMENT GREEK [8116 


might be taken as a sort of double perfective, a combination of διελέγ- 
χομαι confute and κατελέγχομαι convict; but Blass is probably right in 
classing it with διαλέγομαι (see below). In διακωλύω Winer would make 
διά express “‘ intercipiendi et arcendi notionem”: he compares ver- 
hindern against hindern, and διακλείειν intercludere. It is perfective 
in any case: Mt 3 is “John was for stopping him altogether.” 
Διασαφέω (common in earlier vernacular Κοινή) raises a difficulty in that 
σαφέω does not seem to have existed, though ἀποσαφέω is found, another 
perfective: it is likely that these two composita were formed at once 
from σαφής (cf. σαφηνίζω), the simplex being skipped. In that case we 
can enter it still as a proper compound verb, with perfective dia: we 
may compare our verb clear up from the Middle English adj. clear 
(nearly a century older than the verb clear in the N.H.D.). Next to 
these exx. of perfective διά we might put some in which the prefix may 
be rendered thoroughly, but there does not seem to be perfective force. 
Διαβεβαιόομαι affirm strongly, διαμαρτύρομαι solemnly declare, διισχυρί- 
Lopar confidently assert, only differ from the simplicia by the equivalent 
of σφόδρα, no achieved result being implied. So διαπονέομαι, διαπορέω, 
διαταράσσω (perturbo). 


There remain the compounds in which 6:a=dis, between or to and fro. 
The force of dis is easily recognised in διαγινώσκω and διακρίνω (where 
the middle distinguish for oneself naturally develops into hesitate by 
stressing the dis-), in διαδίδωμι, διανέμω, διαιρέω, διαμερίζω, διαρρήσσω, 
διασπάω, διαχωρίζω, διαρπάζω, διασπείρω, διασκορπίξω, most of which 
are perfectives. So also Siadépw=differo, διαστρέφω distort, διατίθεμαι 
=dispono, Svacté\\w=discrimina facere (whence to give express com- 
mands), διατάσσω (sim.), διαλύω ---αὐδϑοῖνο, διικνέομαι penetrate between, 
διίστημι put asunder (διαστῆναι intervene or depart), διανοίγω (Winer 
compares with Ac 75 Virgil’s video medium discedere caelum). In διεν- 
θυμέομαι we think of “ This way and that dividing the swift mind.” 
Διαχειρίζω administer (possibly coloured by the common business phrase 
διὰ χειρός) Seems to have been developed into a euphemism for kill (as 
our settle is sometimes, or despatch). In the rest the mutual relation of 
the A and the B is indicated by mediating διά, as we might call it: 
διαλέγομαι, διαλαλέω, διαλογίζομαι, Svepwrdw recall the two parties in 
a conversation, διανεύω the same for a dumb show, 81axdevdlw perhaps 
is similarly conditioned by the flinging of ridicule at another party. So 
διαλλάσσω is to make a (favourable) change in people who are at variance 
(διαμαχόμενοι) : διαλύω is similarly used in papyri; διαδέχομαι applies 
to one official who succeeds another. 

(b) Some 20 noun compounds of διά are connected with verbs accounted 
for in (a), and the prefix is similarly explained. Sometimes, however, 
the noun happens to fall into a different class, attaching itself to a use of 
the verb which does not occur in NT. Thus διαστολή difference has the 
dis still very marked, which has become latent in διαστέλλω : the verb 


§ 116] WORD-COMPOSITION—PREPOSITIONS. 303 


has the corresponding force in Polybius. Atatayy has the mediating 
διά, not seen in the verb. In διαυγής transparent (cf. duapavjs—printed 
in Rev 217! TR without any authority) διά is much clearer than in 
διαυγάζω. Διάλεκτος ---ἡ διάλεκτος γλῶττα, with verbal adj. in two 
terminations: διά expresses very well the language of ordinary 
intercourse. 

Passing to nouns without congeners in (a), we must pause on the 
old crux διάκονος (“-véw, “via). It is curious that the latest scientific 
opinion has been swinging back to the recognition of this as a διά com- 
pound after all, with ἃ (Ion. διήκονος) affected by διηνεκής (¢.v.) and 
its class. (See Brugmann ap. Boisacq s.v.). In that case we can connect 
with the older word ἐγκονέω (Homer, in pte.) and recognise a metaphor 
from the games (cf. ἀκονιτί ete.): the starting-point would be ἐν κόνϊ 
as an imperative, διακονέω (which would thus come before its noun) 
being like this a compound of Class VII. (above, p. 291). I set this down 
with some reserve. Διηνεκής may be taken next, as in any case con- 
taining é:a=per. Thumb observes (Dial. 117) that the ἡ is Doric as 
well as Hellenistic: Attic διᾶνεκής will thus (like διάκονος ) be due to 
the special Attic sound-law that made ἡ pure into a. Like ποδηνεκής 
reaching to the feet, δουρηνεκής reaching a spear-throw, this comes from 
 enek which makes part of the system of φρέω (cf. the two roots combined 
in our br-ing). The ἡ is from the rule that lengthened the vowel at the 
junction, in compounds whose second element had vowel anlaut. Διάδημα 
goes with διαζώννυμαι : the dual character of διά is dropped like that of 
ἀμφί in ἀμφιέννυμι (p. 294). Διέξοδος if taken according to RV will 
have dua=dis—the streets that come out of the town and fork there. 
See Vocabulary s.v. Διόρθωμα and σις are from διορθόω to correct 
right through (constantly used of what we should call proof-reading). 
Avatpopy goes with the perfective verb διατρέφω (per temporal), but 
the perfective sense is hardly visible. On the other hand, διαπαρατριβή 
(ἅ.λ. in 1 Tim 6°) is a perfective from παρατριβή collision (Polybius), with 
temporal per. Finally comes διανόημα from διανοέω, like διενθυμέομαι 
above, and διάνοια : the formation of this last is probably influenced 
by ἄνοια, δύσνοια, ἔννοια» where there were adjectives with possessive 
force (6... avoos) to start the abstract. 


(c) Διόπερ and διότι are the only words in this class, and of course 
they are only conjunctions formed out of phrases—é.’ ὅπερ and δι᾿ ὅτι, 
for which reason—and not compounds at all. Διότι has come in Hellen- 
istic to be often a mere synonym of ὅτι that, used by Polybius, for 
example, in order to escape hiatus : in NT =because or for. 


1[See, however, the new LS s.v. Mr. E. E. Genner points out that the 
“* Attic”? form only occurs in the Hippias Major (twice on one page), where 
the MSS are not unanimous, whilst διηνεκής occurs in the Laws, where they 
are unanimous.—HD. | 


304 A GRAMMAR OF NEW TESTAMENT GREEK. [§ 117-118 


δ 117. 6. tEis eto is for évs, which still 
survived in Cretan before vowels, its sandhi- 
form ἐς coming before consonants: in Hellenistic ἐς was 
obsolete, only appearing in NT in ἔσω and ἔσοπτρον. It should 
not be treated separately from év, of which it is a variant found 
in Greek alone, being formed on the analogy of é§ for use with 
verbs of motion. Some dialects (Boeotian, Thessalian and 
N.W. Greek—see Thumb Dial. 55, 184) used only ἐν. Its 
compounds are naturally few by comparison, though they go 
back to Homer: zneo is more original than εἴσειμι, infero than 


εἰσφέρω. 


Εἰς 


(a) Ten εἰς composita are left in NT. Going into (leap, run) accounts 
for half of them. Etodyw and εἰσφέρω bring into, εἰσκαλέομαι invite 
into, are all simple. In εἰσακούω alone has there been any obscuration : 
as early as Homer the eis developed the simplex hear into attend to. 
It is not unlike εἰσδέχομαι receive into (one’s house), welcome. Nine of 
these verbs only occur 50 times in NT all told, and εἰσέρχομαι (191 times) 
is the only common one. 

(6) Two compounds of εἰσάγω, ἐπεισαγωγή and παρείσακτος, give 
nothing to discuss under the εἰς heading. Εἴσοδος incoming (in various 
senses) is often in Hellenistic associated with ἔξοδος : the former goes 
back to Homer, the latter to Herodotus and Attic. ὌἜσοπτρον, always 
so spelt, illustrates the fact that the sandhi distinction of εἰς and ἐς was 
lost early (from Pindar down) in this word: on ¥ ὁπ see Prol. 111, and 
on the suffix -rpov below, p. 369. 


(c) Ἔσω (whence ἐσώτερος and ἔσωθεν) is dealt with below, p. 330. 


§ 118. 7. {Ἐν within has been mentioned 
under eis. It is found with the same meaning 
in Lat. in, Goth. in, and in Keltic and Slavonic, etc. Greek 
has besides ἐνς (eis) the form évi, but not (like Skt., Lat., Keltic 
and Germanic) the derivative enter (nétér). It is possible that 
the reduced form » (found in Lithuanian) may account for 
some compounds with initial a-: thus atevns (whence ἀτενίζω) 
suggests intendo more naturally than the prefix sm (ἀκολουθέω, 
etc.). See above, under a-, p. 285. Ἔν is by far the commonest 
of prepositions in NT (Prol. 98), but in forming compounds falls 
behind ἀνά, which is nearly extinct as a preposition. This 
illustrates very well the independent development of the two 
uses of these adverbs which we call prepositions: the facts 


Ἔν 


§ 118] WORD-COMPOSITION—PREPOSITIONS. 305 


presented in this chapter make it easy to understand how pre- 


positions which one language uses largely are only adverbs in 
another. 


(a) Fifty-five composita occur in the NT, which seem about equally 
divided between the forces of eis and ἐν. It will not be necessary to 
mention many of them, as in most the choice between into (on) and in is 
easy, or may be left indeterminate. Those that need a note may be 
taken as we come to them. ᾿Ἐγκαλέω does not intrinsically differ from 
εἰσκ.; but one may “call in” a person for various purposes, and ἐγκ. was 
early specialised in malam partem, calling a man in to accuse him. In all 
languages doublets tend to be utilised for different meanings. So 
€vdexonar=admit to oneself (with ἀνένδεκτος) : εἰσδέχομαι and εἰσκαλέω, 
as is natural in compounds of the more sharply defined εἰς, have mean~- 
ings coming directly from their constituents without later develop- 
ment. ᾿Ἐγκαταλείπω will serve as a type of some others: καταλείπω 
abandon (perfective) is supplemented with ἐν, pointing to the plight in 
which the victim is left. ᾿Ἐγκομβόομαι is not a quite certain com- 
pound, as its simplex does not occur in extant literature: we depend 
on Suidas citing Epicharmus and a later comedian. It is presumably 
to gird something on oneself. ᾿Ἐμβριμάομαι seems to connote strong 
feeling within oneself. “Epmattw=illudo, ἐμπίπλημι =impleo, ἐμπλέκω -- 
implico. Ἐμπορεύομαι is in form a compound, but according to its 
meaning has nothing to do with πορεύομαι: it would indeed be more 
correct to exclude it from the list of composita. Its meaning is entirely ἡ 
determined by ἔμπορος (see below), but had there been no πορεύομαι, 
the verb would more probably have been ἐμπορέω: on this ground we 
may keep it here. The same may perhaps be said of ἐνδοξάζομαι, 
which I suspect was partly influenced by ἔνδοξος --δοξάξω extol is common 
in Hellenistic but very rare in classical Greek. Probably both forces 
acted, for the compound is late: the ἐν has distinct force in one of its 
occurrences at least. “Evopkifw likewise owes its ἐν not so much to signi- 
ficant composition (ἐν + 6pki{w) as to the association with ἔνορκος ‘uratus 
(ἐν ὅρκῳ ὦν, Cl. VIL). “Evdpxopor is supposed by Lightfoot to retain 
a trace of its classical connotation, the initial act of a sacrifice. But in 
Polybius and LXX it is simply begin (qs. make a start wpon an act), and 
it does not seem that the technical force survived. In its two Pauline 
occurrences, as in those of its cpd. mpoev., it is directly contrasted with 
ἐπιτελεῖν (perfective), which marks the last step as the other marks the 
first. ᾿Ἐνδείκνυμαι (cf. indico) suggests completer demonstration than 
the simplex—laying the ‘‘ index” finger, as it were, on the object. With 
ἐνδύνω (ἐνδύω is not found in NT) we may couple ἐνδιδύσκω, an alter- 
native present stem. The ἐν is simple, but the semasiology is hard. The 
meanings clothe and sink are not easy to correlate, and both of them are 
found in Skt., in separate roots which can equally well answer to the 


306 A GRAMMAR OF NEW: TESTAMENT GREEK. _ [§ 118 


Greek: 1 we may perhaps make di-i6 aor. édim (Skt. wpd-du) the I.E. 
word for clothe, and dusné (cf. Skt. dosd “‘ in the evening”’) sink down, set, 
penetrate, whence δυστμή. The two must have been confused in Greek 
from the first. With ἔνειμι inswm we join ἔνι (common in Hellenistic), 
which is simply the alternative form of ἐν with the substantive verb 
understood and meaning specialised.? “Evéxw has the ἐν rather obscured 
by development. When=entangle, hold in, the prefix is clear (cf ἔνοχος): 
the three Biblical passages (so far unparalleled) for évéxw c. dat. =press 
on, set wpon, may be most simply explained by Hesychius’ ἔγκειται, with 
intrans. use of ἔχω and ¢v=against—others assume an ellipse (χύλον or 
simply νοῦν). Ἐνίστημι (only intr. tenses) answers to insto (ev practi- 
eally=cis). ᾿Ἐγκαινίζω innouo (P Par 1633 (B.c. 127), as well as LXX). 
᾿Εγκόπτω must be the opposite of προκόπτω, which is perhaps a metaphor 
from path-making, cut one’s way forward: this will be cut into the path, 
hinder. “Eyxpivw=judge or reckon among. Ἑνοχλέω is probably 
bring annoyance wpon. ᾿Ἐντέλλομαι is difficult, the relation between 
simplex and cpd. being far from clear: moreover, the associated nouns 
ἔνταλμα and ἐντολή, by their form clearly primitive, have no uncom- 
pounded form at all. Unless we make the influence of ἐντέλλω on them 
rather improbably strong, the root should be I.E. tel, not q%el, from 
which other uses of τέλλω certainly come: can we compare intuli, and 
make évté\Nopat—=bring one’s influence to bear upon aman? Ἐντρέπομαι 
is apparently turn towards (in bonam partem), so regard. ᾿Εντυγχάνω 
(whence in a special sense ἔντευξις) fall in with, meet. 


(6) The main points requiring settlement under this heading concern 
the recognition of compounds belonging to Class VII. Nearly a score 
of nouns and adjectives attach themselves to verbs occurring in NT, and 
either explained above or omitted as self-evident. Nearly all the rest, 
however, will need some comment. “Eyxd@etos (from ἐγκαθίημι let down 
tnto)=suborned. “Eykpatys (Tea “τεύομαι) --λαυΐηρ strength within, 
self-controlled.2 ᾿Ἔγκυος (κύος) having’ fetus within. ᾿Ἐμβατεύω," 
which in the absence of any βατεύω must come in this class, raises some 
difficulty, in that ἐμβάτης, its presumed noun original, begins to appear 
in literature nearly a century later than the verb, and then in a sense 
which cannot help us. But ἐπεμβάτης, mounted, is classical, and so are 
ἐπιβάτης and ἐπιβατεύω, which between them probably formed ἐμβατεύω 
directly: the sense in Hellenistic is always enter wpon (also class.), of 


1 Latin ind-vo and ex-uo can have no etymological connexion with ἐνδύω, 
for all the similarity of form and meaning. It is possible, however, that induo 
may be a conflation of ind-euo (endo, our into) and in-duo=évétw. 

2 Its development in medieval Greek may be seen in Dieterich Unters. 
225 ff. The MGr eivac=éori is ἔνι with the vowels assimilated to those of 
εἶμαι and εἶσαι (sum, es). 

5. But its opposite ἀκρατής suggests the possibility of its being a Cl. VII. 
epd.—see below. 

4 Very often ἐμβαδεύω in papyri: cf. Wilcken, Ostr. i. 190 f. 


§ 118] WORD-COMPOSITION—PREPOSITIONS. 307 


taking possession. (See also above, p. 273, on κενεμβατεύω.) ᾿Εμφανής 
(whence “vifw) resembles ἐναργής, which differs from apyjs apparently 
in the suggestion of inherent light that makes the compound more 
emphatic: the Greek sense of the word we have just used is another case 
in point—ef. ἔμφασις in LS 5.0. Π1. ᾿Ενδεής having want within presumes 
(like ἐπιδευής, ὑποδεής) a neuter noun *SéFos want. ᾿Ἐνδώμησις (so N*A 
in Rev 21!8: see p. 73, and WH App.” 159)! building in, from δωμάω 
(./ dem with long grade). “Evédpa (whence “petw) sitting within (in hiding), 
exactly as the independent Lat. insidiae. Ἔννοια is apparently from 
ἔννους (see on διάνοια above) having mind within, intelligent. But as 
the opposite of ἄνους, we might as well put ἔννους below. “Evtadidtw 
(whence “opos) is from ἐντάφιος funereal, which is most naturally ex- 
plained as an adj. from the root of ἐν θάπτω bury in: Class VII., however, 
(ra ἐντάφια obsequies=ra ἐν ταφῇ), is quite possible, though less probable. 
Παρεμβολή army, camp, or barracks, seems originally to have meant troops 
drawn up for battle, cf. παρεμβάλλω. This specialisation of meaning is 
said to have started in Macedonia. In Attic at any rate an entirely 
different use appears, insertion, depending on ἐν in, while the other is 
linked with ἐμβάλλω attack (¢v=against). 

(c) There is a considerable proportion of Class VII. compounds made 
with ἐν. Indeed, it is quite an open question whether a good many 
of the compounds classified above may not have originated rather like 
those to be described now. Compounds of ἐν are largely found as anti- 
thetic to those in a- privative, which have the meaning without . . . (lit. 
having no... ). So ἄνομος without law is opposed to ἔννομος under 
law in 1 Co 971, ἄτιμος without honour to ἔνδοξος with glory in 1 Co 410 
(ἔντιμος elsewhere). These must be interpreted as adjectives directly 
formed from ἐν νόμῳ, ἐν δόξη, ἐν τιμῇ. So ἔμφοβος (opp. ἄφοβος), 
ἐνάλιος, ἔνδημος (whence “pew, opp. ἀπόδημος), ἔνδικος (opp. ἄδικος), 
ἔννυχος (only adv.—the x for κττεχτ represents a simpler form of the 
root), ἔντρομος, ἐντόπιος (with further suffix, like ἐνάλιος), and so 
ἐνύπνιον (ἐνύπνιος from évumvos). Some others need detailed treat- 
ment. Ἔγγυος goes with éyyin security, which is probably from an old 
word for hand (cf. yiadov, Lat. vola, Av. gava-), like ἀμφίγυος and 
ὑπόγυος, So=what is putinthe hand. “Eykaxéw is best taken as ἐν κακῷ 
εἰμί “ἴῃ a bad way,” ill, enfeebled: this depends mostly on a physical 
connotation of κακός, while ἄκακος depends on the moral, and so does 
ἐγκακέω itself in Polybius (iv. 19. 10). ᾿Εγκρατής (ete.—see under δ) 
might be reckoned here because of its opposite ἀκρατής without self- 
control. Since κράτος suggests the possession of strength, as βία the 
using of it, ἐν κράτει (év)=self-controlled involves only the specialising 
reference to moral κράτος. ᾿Ἑλλογάω (or -€w—see §§ 84-5)=ev λόγῳ 
τίθημι, according to the very common use of Adyos=accounts: so im- 
puto, put to the account of. There is no connexion with ἔλλογος rational 


1 Syll. 583%° (= Syll.3 996%) (i/a.p. ἢ), BCH xvii. 78 (no. 1°). 


308 A GRAMMAR OF NEW TESTAMENT GREEK. [§ 118-119 


(cf. ἔννους), the opposite of ἄλογος, from ἐν λύγῳ in the other sense. 
‘Evavtios (and rev.) is probably a cpd. of ἀντίος (p. 296). But while 
ἔναντι and such compounds (p. 329) are correlative adverbs whose con- 
tents are added together (“‘ within opposite,” etc.), the similar-seeming 
adj. must be explained differently, though in this case the closely parallel 
compounds affected one another. "Evaytios=ev ἀντίῳ (dv). For the 
relations of this group see Wackernagel Hellenistica pp. 1 ff. ᾿Ενθυμέομαι 
(-novs) comes directly from ἐν θυμῷ (ἔχω) : the negative ἄθυμος (whence 
“μέω) starts from a different association of θυμός, with εὔθυμος as anti- 
thesis. Ἔνεργός (whence “yéw and its derivative Ὕημα) appears to be 
from ἐν ἔργῳ active, or productive, as opposite of ἀργός inactive or barren. 
The formation of its later equivalent ἐνεργής is not perfectly clear: the 
derivative ἐνέργεια speaks for an original -es- stem, but its record is not 
good enough to claim antiquity. Has it been formally assimilated to its 
like-sounding adj. ἐναργής clear? On the whole group see J. A. Robin- 
son’s important excursus, Hphes. pp. 241-7. ᾿᾽Ἐνιαυτός year, originally 
anniversary, is taken by Prellwitz as starting from ἐνὶ αὐτῷ (sc. I suppose 
ἤματι, ““on the same day”). But see Brugmann Gr.4 195 n.1, and below, 
p. 372. Ἐνωτίζομαι --ἐν dri (ἔχω) is formed directly, in complete 
independence of ἐνώτιον earring (Attic év@Svov—Meisterhans? 79). 

Apart from the cpds. named in ὃ 130, the following adverbs are con- 
nected with ἐν. “Ev@d8e here or hither (the two being confused in Hellen- 
istic) is from ἔνθα, with suffix - θα, which may be compared with Skt. μάλα 
(where 1), thd (here): Brugmann KVG 455 gives this doubtfully, and (7b. 
456) denies the equation ¢v@a=inde (as far as the in- goes). The suffix 
-de=Eng. to in origin and meaning. Ἔνθεν has the ablative suffix -θεν. 
᾿Εντεῦθεν has suffered a shifting of aspiration in Attic: Ionic ἐνθεῦτεν is 
original. Brugmann Demonstr. 104 n. shows how the flexion of οὗτος 
produced a whole series by analogy—éevécirey came from ἔνθεν as 
τηλικαῦτα from ταῦτα ete. 


Ἐξ δ. 119. 8. ἘἘξ, ἐκ, owt. The primitive form 
was eghs, as Brugmann (KVG@ 179) shows from 
Locrian ἐχθός (Att. ἐκτός by anal. of ἐντός) and ἔσχατος (from 
eghs-qo- becoming egsgho-): the by-form ἐκ (also ἐγ) is due to 
the dropping of o between two mutes (see Brugmann Gr.* 148 f.). 
Latin and Gaulish ex are among the cognates, which confirm the 
meaning given: in the Keltic branch ez (Ir. ess) is only used in 
compounds, which illustrates its large proverbial use in Greek. 


(a) Ninety-four composita with ἐξ are found in NT. We can re- 
cognise owt in about fifty of them very clearly, and in many more with 
an easy adaptation: naturally the prefix has, like its English equivalent, 
a considerable range of meaning. In cpds. where the local force 
is marked, there is nearly always perfective action: indeed, it is only 


§ 119] WORD-COMPOSITION—PREPOSITIONS. 309 


when owt is weakened, and from, forth or off gives the sense, that we have 
exceptions. Thus ἐκκρέμομαι hang from, i.e. wpon: ἐκτείνω stretch out 
and ἐκπετάννυμι spread out are the nearest NT exx. to go with it, but 
even here ἐκ shows the action of the verb carried as far as it will go, so 
that we may class both verbs as perfective. The simplest local sense is 
seen (as with ἀπό etc.) in verbs of going—egerpr and ἐξέρχομαι, ἐκβαίνω, 
ἐκπορεύομαι etc., or allied conceptions, as ἐκπέμπω, ἐκβάλλω, ἐκκολυμβάω, 
ἐκκομίζομαι, ἐκπηδάω and ἐξάλλομαι, ἐκπλέω, exowlw (Ac 2789), 
ἐκτινάσσω, ἐκφεύγω, ἐξανίστημι and ἐξεγείρω, ἐξαιρέω, ἐκκλάω, ἐκκόπτω, 
ἐξορύσσω, ἐκχέω (-χύννω). ἐξωθέω etc., involving removal owt of one 
place into another. Local force of slightly varying kinds (literal 
or metaphorical) is seen in ἐξαγγέλλω and ἐξομολογέω;; ἐκλαλέω 
let out, divulge; ἐκνήφω of sobriety attained out of drunkenness; 
ἐκπνέω and ἐκψύχω (strongly perfective); ἐξαιτέομαι beg surrender 
of (removal from present position), ἐξαγοράζομαι ransom (from bondage 1) ; 
ἐκμάσσω wipe out and ἐξαλείφω smear out, ἐκκαθαίρω clear out; ἐκνεύω 
(perhaps a metaphor from driving a horse); ἐξίστημι (orig. displace— 
ellipsis of φρενῶν or the like accounts for specialised meaning) ; ἐκδίδομαι 
let out (but act. in sense surrender, in the verbal ἔκδοτος) ; ἐκκαίω (of 
fire blazing ouf—the perfective force (metaph.) is very clear in Rom 157 as 
compared with the simplex of the same metaphor in 1 Co 179), ἐκλάμπω 
to shine out; ἐκπορνεύω ; ἐκλείπω is primarily “ leave off’; ἐξορκίξ 
(verb purely perfective in NT occurrence =adjure, bind by an oath, but 
elsewhere like its noun ἐξορκιστής exorcise, get an evil spirit owt by potent 
words) ; ἐκτίθημι (ἔκθετος) as in Ac 719: 22; ἐκλεγομαι eligo, choose out 
of a larger number; ἔξεστι (ἐξουσία, “alw) it is out, i.e. open, allowed, 
which comes curiously near its apparent opposite ἔνεστι in its classical 
meaning τ is possible. 

In a very few cases ἐξ may be rather intensive (out and out) than 
perfective. ᾿ΕἘκθαμβέομαι and ἐκθαυμάζω might be described as per- 
fective in that they denote a complete astonishment, but a perfective 
does not differ from a non-perfective merely by the addition of very. 
᾿Εκμυκτηρίζω primarily may be only “scoff bitterly,’ but might be 
called perfective as much as καταγελάω laugh to scorn. “Extapdcow is 
graviter turbare—but conturbare (Vulg.) is perfective. ᾿Ἐκπειράξω (τὸν 
θεόν or equiv.”) might suggest the daring of the act, or we might find 
in it the effort to put to a decisive test. It would not be difficult, indeed, 
to trace in all of these a nuance that would bring them among the per- 
fectives. ᾿Εκδύω may be noted here as a verb which, though perfective 


1So J. A. Robinson even in Eph 5'® (see note there), Col 45=redeeming 
what has fallen into bad hands. For other views see Peake and Lightfoot on 
Col. Lc. 

2 In Dt 85:16 God putting Israel to a thorough test. This may be the mean- 
ing in Lk 1055, or the Evangelist may be using instinctively of Jesus a word 
normally used of God. 


310 A GRAMMAR OF NEW ‘TESTAMENT GREEK. [8110 


already, forms a further perfective : it might imply only the putting off of 
certain garments, while ἀπεκδύομαι, like its noun ἀπέκδυσις, connotes 
complete stripping, of oneself or another in one’s own interest (Col. 
gill. 15 3°). 

The following are the perfectives in which the local force of ἐξ has 
wholly or nearly disappeared. We take first those on which some com- 
ment is wanted. 2 Co 4° ἀπορούμενοι ἀλλ᾽ οὐκ ἐξαπορούμενοι, per- 
plexed, yet not unto despair, is a specially good example of perfective 
action. “Ex{ytéw always seems to denote that the seeker finds, or at 
least exhausts his powers of seeking (Heb 1217): so ἐξεραυνάω, ἐξετάζω 
(rare simplex ἑτάζω--ὦ in Arcadian, suiting its probable connexion with 
J es to be—akin to reds and grupos, so verify). The Hellenistic verbs 
ἐξουθενέω and ἐξουδενέω (NT and BGU iv. 1117* (13 B.c.—reading not 
certain)) are good exx. of the continued vitality of this word-forming 
process. Thackeray shows (Gr. 105) that the former was coined first, 
during the vogue of οὐθείς (see p. 111), and ἐξουδενόω later, when οὐδείς 
had begun to reassert itself: ἐξουδενέω is due to mixture. Independent 
of both is Plutarch’s ἐξουδενίζω (see above, ὃ 46): the forming of com- 
pounds on the basis of the neuter οὐδέν goes back to Plato’s οὐδένεια. 
We have then three separate verbs for ‘“‘ making nothing of, despising,” 
formed from ἐξ and the word for nihil without intervention of a simplex 
verb. Probably ἐξ was appropriate not only for the needed perfective 
force, but also to make the transitive clear—a function these prefixes 
often tend to achieve. ᾿Ἐκδέχομαι in its NT sense (so Sophocles and 
Herodotus) is a little obscure. Jebb (on Philoct. 123 σὺ μὲν μένων νυν 
κεῖνον ἐνθάδ᾽ ἐκδέχου) notes: “‘ é., excipe. The idea of the compound 
is, “ be ready for him,’—prepared to deal with him the moment that he 
appears.” This fits the perfective idea when we remember the present- 
stem action (Prol. 114): qs. ‘‘ go on with the act of ‘ receiving’ till he 
comes.” (We may compare the way in which μέλλειν to be about to 
gets from the durative present the sense of delaying.) ᾿Εξηγέομαι is 
in NT always “ explain,” se¢ forth, as already in (post-Homeric) classical 
Greek. The comparison of cognate words (Lat. sdgio, sdgaz, Goth. 
sokyan, Eng. seek, Ir. saigim, with same meaning) shows that the meta- 
phorical application in ἡγέομαι was there from the first: ἡγεῖσθαί τινι 
would be explore for, blaze a path for. ᾿Ἑξηγεῖσθαι then is the perfective : 
ef. ἐκξητέω (above) and ἐκθέσθαι exponere. ᾿Ἐξισχύω in Eph 318 c.inf. is 
a striking perfective—‘‘ be strong enough” to apprehend, a strength 
exerted till its object is attained. “Expi{dw is in form a compound, and 
of course perfective. But the meaning wproot cannot be deduced from a 
combination of ἐκ and ῥιζόω, and we must explain it as we explained 
ἐμπορεύομαι above (p. 305): its meaning is determined by ἐκ ῥιζῶν 
(ἀναιρεῖν), just as eradicare is really short for ex radicibus evellere. It is 
therefore virtually a cpd. of Class VII. Most of the other perfectives 


1 So (virtually) first AV. The earlier vss. take d7.=be poor, but represent 
the antithesis correctly. 


ere oan 


§ 119-120] WORD-COMPOSITION—PREPOSITIONS. 911 


need only be named. “ExSamavdw spend out, spend wholly ; ἐκδιηγέομαι 
tell right through to the end; éxxevtéw to give a deadly stab; ἐκλανθάνομαι 
quite forget ; ἐκλύομαι faint off (lit. be dissolved outright) ; ἐκπληρόω fill 
up, as we say ; ἐκπλήσσομαι, much like ἐκφοβέω, and ἐκταράσσω above— 
our use of strike for a mental impression will illustrate ; éxtehéw (cf. 
ἐξαρτίζω) finish off ; ἐκτρέφω rear wp, in our idiom, decidedly perfective— 
nourisheth is too weak in Eph 5°, and nurture (RV) in 64; ἐξακολουθέω 
follow out; ἐξαπατάω of successful deceit!; ἐξολοθρεύομαι destroy 
utterly. 

(6) There are some twenty-five noun cpds. which attach themselves 
to verbs accounted for in (a), and only a few remain. Local ἐξ is re- 
cognisable in διέξοδος (see διά), ἔκγονος, ἐκκλησία (orig. a summoned 
assembly), ἔκτρωμα (ἐκτιτρώσκω), ἐξέραμα, ἔξοδος, ἐξοχή (ἐξέχω) of 
outstanding prominence. Perfective €§ appears in ἀνεξιχνίαστος (cf. 
the ἐκζητέω class above), ἔκδηλος, ἔκτρομος (only ND in Heb 12%— 
probably assimilated to ἔκφοβος in context, on which cf. ἐκφοβέω). 

(c) To Class VII. belong the following. Ἔκδικος in earlier Greek is 
ἐκ (1.6. ἄνευ) δίκης, ealex, but as early as the Batrachomyomachia shows 
the distinct meaning avenger: see Milligan Thess. 50. This may possibly 
be derived from another use of ἐκ, seen in the recurrent formula in legal 
papyri, καθάπερ ἐγ δίκης ‘‘ just as after a legal decision” : ἔκδικος would 
thus be one who carries out asentence. There is, however, the alternative 
possibility that the classical compositum ἐκ δικάζω avenge has influenced 
the meaning of ἔκδικος (with its deriv. “Kkéw “κημα). Ἔξυπνος (hence - 
“νίζω) --ἐξ ὕπνου (γενόμενος). Συνέκδημος (see σύν) is a compound of 
ἔκδημος -εἐκ δήμου (dv), like ἀπόδημος. 

Perfective ἐξ appears strengthening an adverb in ἐκπερισσῶς (ὑπερεκπ. 
1 Th 53 BDG—see ὑπέρ). Ἔκπαλαι is like ἀπὸ τότε, no real compound : 
cf. ἐξαυτῆς (really two words). ᾿Εξάπινα (earlier ἐξαπίνης, with ending 
assimilated to adverbs) and ἐξαίφνης are no doubt similar phrases, but 
the second element is not clear: αἶψα is likely enough for the second 
word, but will not suitthe first. For ἔξω and its derivatives and ἐκτός, 
see § 130. 


§ 120. 9. ἜΠ  Επί (independent accent form 
ἔπι), near, on, wp to, appears in three gradation 
forms, answering to I.E. opi (ὄπι-θεν, Lat. op-, ob), epi (ἔπι)--- 
Skt. dpi, Av. a*pi may be either ; and pi (πιέξζω Skt. prdayati= 
pi-2zd- »/ sed (sedére); also wt-vy whence ἀνα-πτύσσω, from 
pitugh, Skt. pyuksna). 

(a) As many as ninety-nine composita with ἐπί occur in NT, a total 
only exceeded by κατά and σύν, the main perfectivising prefixes. Accord- 


’ 
Ἐπί 


1 Conative force may be added in present stem—see Prol. 114. 


312 A GRAMMAR OF NEW TESTAMENT GREEK. [8 120 


ing to some scholars, a contributory cause for its great frequency in all 
periods of the language was its inheriting much that belonged to another 
preposition, akin to Skt. abhi, Zd. aiwi, Eng. by. The hypothesis explains 
one or two forms like ἐφίορκος, and Delbriick brings up a good many 
parallels between ἐπί compounds and Skt. abhi- compounds: there is 
also the advantage of a slightly easier explanation of the range of mean- 
ing found in the ἐπί family. But it cannot be said that the phonetic 
argument for the conflation has much weight ; and the hypothesis, which 
is opposed by Brugmann, must remain only a possible one at best. We 
may still observe (with Brugmann KVG 467) that in Latin ob has taken 
over some of the functions of obhi: bhi, though actually derived from 
opi ; and similar mixtures may have occurred in isolated cases elsewhere 
without demanding the fusion postulated by Delbriick. 

In many of the verbal ἐπί cpds. we can trace a clear local sense, and 
these are clearly the oldest. Delbriick (Grd. m1. (i.) 675) gives some 
which may have been compounded before the ethnic period. Closely akin 
to these are the composita in which the preposition may be described as 
directive, indicating the concentration of the verb’s action upon some 
object: in these cases the simplex will be general and the compositum 
special in its force, the one may be abstract and the other concrete. The 
exx. which Dean Robinson chooses for his “ directive ’’ sense are “ ἐπαινέω, 
ἐπιδείκνυμι, ἐπιζητέω, ἐπικαλέω, ἐπικηρύσσω, ἐπικρατέω, ἐπικρύπτω, 
ἐπιμέλομαι, ἐπιμιμνήσκομαι, ἐπινοέω (excogitare), ἐπιχορηγέω," as well 
as ἐπιγινώσκω, the special subject of his well-known excursus (EHphes. 
248 ff.). We may add to the class ἐπαγγέλλομαι, ἐπαγωνίζομαι, 
ἐπαθροίζομαι, ἐπαίρω, ἐπαισχύνομαι, ἐπαιτέω, ἐπακολουθέω, ἐπακούω 
and ἐπακροάομαι, ἐπανάγω and the other cpds. in ἐπ- ανα- (with hostility 
implied in ἐπανίσταμαι), ἐπαρκέω, ἐπαφρίζω, ἐπεγείρω (hostile), ἐπεῖδον, 
ἐπεισέρχομαι, ἐπεκτείνομαι, ἐπερωτάω, ἐπέχω (in some forces), ἐπιβάλλω 
(or local), ἐπιβαρέω, ἐπιβλέπω and ἐπισκέπτομαι and ἐπισκοπέω, 
ἐπιγίνομαι, ἐπιδέχομαι, ἐπιδίδωμι, ἐπιδύω, ἐπικαλύπτω, ἐπίκειμαι (or 
local), ἐπικέλλω, ἐπικρίνω, ἐπιλαμβάνομαι, ἐπιλανθάνομαι, ἐπιλέγω and 
ἐπονομάζω, ἐπιλέγομαι (choose), ἐπιλείπω, ἐπιλείχω, ἐπιλύω, ἐπιμαρτυρέω, 
ἐπιμένω, ἐπινεύω, ἐπιπλήσσω, ἐπιποθέω, ἐπισκευάζομαι, ἐπιστέλλω, 
ἐπιστηρίζω, ἐπιστρέφω, ἐπισυνάγω, ἐπισυντρέχω, ἐπισχύω, ἐπισωρεύω, 
ἐπιτάσσω, ἐπιτελέω, ἐπιτρέπω, ἐπιτυγχάνω, ἐπιφαίνω and ἐπιφαύσκω 
and ἐπιφώσκω, ἐπιφέρω, ἐπιφωνέω, ἐπιχρίω. 

Two more members of this class require a special note. ᾿Επίσταμαι 
is an old compound not very easily analysed: it seems to be ém-oTdpat 
(not duplicated—cf. φημί, ἐφάμην) with meaning “ put oneself in position 
for.” Our understand, Ger. verstanden, will show that the root std is 
capable of application to the mind. It was no longer felt to be a com- 
pound verb. ᾿Ἐπιτιμάω comes from a meaning of the simplex not 
found in NT=/lay penalty on, and so censure. In many of these, which 


1 Especially Delbriick Grd. τη. (i.) 675 f., 679. 


§ 120] WORD-COMPOSITION—PREPOSITIONS. 313 


account for some three-quarters of the total, the line is hard to draw as 
against the purely local force, which I have restricted almost entirely to 
verbs of motion with éri=to or upon. It will be seen at once that the 
directive force as understood here is much wider than is implied in the 
exx. given by Robinson. We might subdivide, but the general nexus is 
clear enough. The only compounds left now are a few in which ἐπί means 
in addition, a natural development of wpon: thus ἐπενδύω, ἐπιγαμβρεύω 
(make a new marriage connexion), ἐπιδιατάσσομαι, ἐπιδιορθόω. 


(Ὁ) About one-third of the noun compounds of ἐπί in the NT attach 
themselves to composita accounted for in (a). Only one of these needs 
separate mention, ἐπιούσιος, the well-known crux in Mt 64=Lk 113. 
That this ἅ.λ. was a coinage by the author of the Greek Q may be safely 
assumed, after Origen. If so, the scholarship of its origin does not justify 
our laying too much stress on considerations which would weigh with us 
if Luke himself or even Matthew were responsible, still less some Atticist 
scholar. Jerome’s swpersubstantialis | is therefore not finally discredited 
by the objection against the non-elision of ει: ἐπιούσιος would thus 
become a Class VII. cpd. rather like ἐπουράνιος (see below), from ἐπί 
upon and so above, and οὐσία. The only meaning quotable for this noun 
from NT and papyri is property or estate, which is not hopeful; and even 
if it is found in an abstract sense in magical documents, this is most im- 
probable in a context so simple and direct as the Lord’s Prayer. Light- 
foot’s argument 2 has not been superseded by later literature (see WS 1. 
136); and we may perhaps agree with Schmiedel in the “ sicherlich ” 
with which he put down ἡ ἐπιοῦσα as the true etymon. ἴἘπειμι (obire 
etymologically) is to come close after, tread on the heels of, as it were ; 
and its participle was current, as the NT itself shows, to express the 
immediately following day. In the evening it would mean the same as 
ἡ αὔριον, but in the morning (see esp. the opening of Plato’s Crito) it is 
a day earlier than this. The immediacy is quite sufficient reason for the 
translator’s being dissatisfied with τὸν τῆς αὔριον or the like as a render- 
ing of the Aramaic before him: he followed a right instinct in coining 
a new adjective from the common word for “ the coming day.” That 
such scholars as Delitzsch and Keim (ap. Lightfoot op. cit. p. 226) should 
have imagined inconsistency with Mt 6** only proves that the succession 
of Martial’s Graeci quibus est nihil negatum is not yet closed. The only 
serious alternative to the above account of ἐπιούσιος is that proposed 
by Debrunner, and epitomised by himself in his Blass Gr.475. He 
makes it a substantivising of ἐπὶ τὴν οὖσαν (ἡμέραν) “ for the current 
day”: for this use of ὧν cf. Prol. 228. He compares ἐπιμήνιος (Poly- 
bius) ‘‘ for the current month,” ἐφημέριος etc. ; but modestly claims only 
a preference for this over the derivation from ἡ ἐπιοῦσα. The lamented 


1 Origen seems to have started the idea. Jerome meant by his word, he 
tells us, swper omnes substantias, extra omnia, and so praecipuus, egregius, 
peculiaris. 

2 Ona Fresh Revision’, p. 217 ff. But see below. 


314 A GRAMMAR OF NEW TESTAMENT GREEK. [§ 120 


Albert Thumb accepted Debrunner’s view whole-heartedly : a pencilled 
note from his hand on this page, which he saw on a visit to me a few months 
before his death, refers to Brugmann-Thumb 675 for his endorsement of 
it. On the other hand, Deissmann' holds to the Lightfoot theory, rightly 
laying stress on the fact that ἡ οὖσα without ἡμέρα expressed has not 
been found. He even suggests that the later ἡμερούσιος, and ἐφ- and 
καθ-ημερούσιος, found in papyri and the anti-Christian writer Proclus, 
are modifications of ἡμερίσιος based on ἐπιούσιος as an “ already 
existing vernacular word”’ used by the evangelists.2 Origen’s ignorance 
of it is met by Deissmann with the remark that he need not have known 
all the popular vocabulary of the Mediterranean littoral. Whether 
this be tenable or not, I should give my own vote for ἡ ἐπιοῦσα with 
much the same degree of preference as Debrunner himself shows in 
voting for ἐπὶ τὴν οὖσαν. On the legitimacy of the hiatus involved in 
Debrunner’s theory, see above, § 38. 

Directive ἐπί may be recognised in the following, with the same 
latitude of application that we found in the verbs :—E7avdp§wots 
(like the composita of ἐπί- ἀνά in (a)), ἐπάρατος and ἐπικατάρατος 
(ἀρᾶσθαι might be cursing at random—the cpd. has an object in view), 
ἔπαυλις (perhaps formed from αὖλις by influence of ἐπαυλίζομαι), 
ἐπεισαγωγή (€reurdyw=introduce as a_ substitute), ἐπήρεια (papp.) 
whence ἐπηρεάζω (dpe) in Homer=violence), ἐπιβουλή, ἐπιεικής and 
ἵκεια (Feixo—pf. FéFoxa—in Homer=suit or agree—cpd. suggests the 
environment which is “‘ suited”’), ἐπιθυμέω “pia “μητής (the stage *éxi- 
Oupos="* having one’s θυμός towards,” is apparently passed over—cf. 
ἐνθυμέομαι above), ἐπίκουρος whence “pla, and “petos from the proper 
name ( ,/ gers=run—cf. curro, horse, and the Keltic original of car), ἐπίνοια 
(see on διάνοια and ἔννοια above), ἐπισιτισμός (from “ἰζομαι, get σιτία 
for definite people), ἐπισφαλής (directing danger to certain objects), 
ἐπιφανής and ‘vera (ἐπιφαίνω has been dealt with, but this comes from 
a technical sense—a divine being manifested to human eyes in human 
form), ἐπόπτης and “τεύω ( ,/ og on =see—cf. ἐφορᾶν, ἐπιβλέπειν etc.), 
ἐφευρετής (from “picka, to find or invent for a purpose). 

Local ἐπί may be seen in érionwos=with σῆμα upon it, the opposite 
of ἄσημος, and in ἐπιγραφή and ἐπίθεσις, the verbs of which belong to (a). 
᾿Επίλοιπος, left over, has the sense added to. The name ᾿Επαφρόδιτος 
(with short form “Etadpas*)=having Aphrodite wpon him: in earlier 
Greek this was venustus (Horace’s Veneris muneribus potens), but later 
represented felix, from the Venus-throw of the dice—so as epithet of 
Sulla. ᾿Ἐπίορκος (“kéw) raises various difficulties. Its frequent appear- 


1 Επιούσιος, in Ntliche Studien Heinrici dargebracht (1914), pp. 115-9. 

2 [Deissmann’s guess finds support by the discovery of the word ἐπιουσί[ων] 
in an old housekeeping book given in Preisigke, Sammelbuch, Nr. 5224. See 
also Deissmann, LAH? 78, τι. 1.—Ep.] 

3 Not implying any necessary identification of the persons. 


§ 120-121] WORD-COMPOSITION—PREPOSITIONS. 315 


ance in the Kown as édiopxos! (Prol. 234) was explained by Thumb 
(Spir. Asp. 72) as due to contamination of ἔφορκος and ἐπίορκος : he 
compares edi ἱερέως on an inscription. Against Osthoffs conjecture— 
accepted by Delbriick—that the I.E. ebhi (Skt. abhi, Eng. by) underlies 
it, there is the fact that ἐφίορκος is late: Phrynichus is right, from his 
point of view, when he bans it as un-Attic, and thereby evidences its pre- 
valence in his own day. Delbriick (Grd. 11. (i.) 676) defines it hesitatingly 
as concerning “‘ the oath by which one presses on a person and compels 
him” ; but this does not explain the false oath. Is against the force of 
ἐπί here, whether drawn from epi or ebhi Ὁ 

(ὁ) The remaining compounds are of Class VII. ᾿Επιτήδειος as a 
difficult word may be mentioned first. Brugmann Dem. 140 ff. suggests 
that ἐπιτηδές, from which it is the adj., starts from ἐπὶ τὸ ἦδος =good 
for that (purpose), τό being demonstrative and ἦδος a rare Homeric 
noun =d¢edos : he compares idoneus (see p. 293 n.1). ᾿Επάναγκες is pre- 
sumably neuter of *é€ravayxys, which might even be ἐπ᾽ ἀνάγκης used as 
an adj. unaltered : in any case it starts from it. (’Emdvayxos occurs also 
in papyri and inscriptions, even as early as the Gortyn Law—-see v. 
Herwerden 8.0.) ”Emapxos (whence °x(€)ia) is from ἐπ᾿ ἀρχῆς, ἐπίγειος 
from ἐπὶ γῆς, ἐπίδημος (whence “péw, παρεπίδημος) from ἐπὶ δήμῳ, 
ἐπιθανάτιος --ἐπὶ θάνατον κείμενος, ἐπικεφάλαιον (Mk 12: D k and 
two important cursives)=70 ἐπὶ κεφαλὴν πρασσόμενον, ἐπουράνιος -- 
ἐπ᾽ οὐρανοῦ ὦν, and ἐφήμερος (pia) is from ἐφ᾽ ἡμέραν. Verbs in this 
class are ἐπιστομίζω from ἐπὶ στύμα (θεῖναι) like the proverb βοῦς ἐπὶ 
γλώσσῃ, and ἐπιχειρέω --χεῖρα θεῖναι ἐπί τι: in neither case does a 
noun compound intervene. Both are ancient words. 

Ἐπάνω --ἐπί-ἀνά- ὦ (see § 130, s.v. ἔξω). ᾿Επαύριον is a true com- 
pound, ἐπί giving the force of succession : its form echoes its primary, the 
adverb αὔριον. ᾿Επέκεινα, to your side, is from ἐπ᾽ ἐκεῖνα. ᾿Εφάπαξ might 
just as well be written as two words, like ἐπὶ τρίς, at any rate for the 
meaning at once (1 Co 15%): when=once for all it is more like a compound, 
a directive strengthening of ἅπαξ in the same sense. 


§ 121. 10. *f{kardé is taken by Brugmann 
(KVG 479) as an extended form of ko or kom, 
found in Latin cum and co-, in Keltic, and in xowds=Kopios : 
Gothic handugs “ wise” (J dhé θη), qs. “capable of mental 
synthesis.” Kdra (so accented as an independent word) is for 
xm-rta, with second element as in pé-ra. Its nearest connexions 
are Irish cét-, Old Welsh cant=along, by, with. Brugmann defines 
its earliest use as “ along something so as to remain in connexion 


Kata 


1 So Mt 533 x, 1 Ti 119 D*P: in LXX three times, in B, A and C respectively 
(Thackeray 126). Ifitis Western (WH App.? 151), it is only another instance 
of Western agreement with the most genuine Κοινή. 

VOLE PART Ti. ——22 


316 A GRAMMAR OF NEW TESTAMENT GREEK. _ [ὃ 121 


and contact with the object.” The kindred Latin contra illus- 
trates one development, agaist, but the meaning down can 
hardly be said to be explained. It is noteworthy that neither 
down nor against survives in MGr.! As the word had of course 
a long history in pre-Homeric Greek beyond our reach, the 
obscurity of its semasiology is not strange. Most probably 
the key should be found in the antithetic development of ava 
and κατά, which with acc. come very near: such antithesis as 
ava ῥόον and κατὰ ῥόον might be very productive. 


(a) Κατά forms 107 composita in the NT, falling thus only a little 
short of σύν, the other great perfectivising preverb. It will be noticed 
that Latin co-, com-, which conspicuously performs this function in Italic, 
is related to κατά in form and to σύν in meaning. To this class belong 
fully two-thirds of the κατά compounds: we cannot always, however, 
classify with confidence, and the ambiguous elements in the word’s 
history cause difficulty. There are some clear instances of local force : 
thus καταβαίνω, καθίημι, κατανεύω and κατασείω (when compared with 
dvavevw), κατάγω, καταπλέω, κατατρέχω, κατέρχομαι. In others this 
meaning could be recognised if the centrality of the meaning down could be 
accepted. Καθέζομαι and κατάκειμαι, for example, are rendered svt 
down, lie down, καταπίπτω fall down, καταπατέω tread down; but we may 
ask pertinently how it is possible to sit, lie, fall or tread in any other 
direction—even if it be allowed that we may “sit up” and even “ lie 
up”! It is better to regard them as simply perfective, the preverb 
(whether down or over or together) bringing the action of the verbal root 
to a definite result. In another class of verbs κατά seems to be repre- 
sented by against, as often when placed with a genitive as a preposition. 
Thus καταβραβεύω, καταγινώσκω, καταγωνίζομαι, καταδυναστεύω, 
κατακρίνω, κατακυριεύω, καταλαλέω, καταμαρτυρέω, καταναρκάω, κατα- 
σκοπέω, καταστρηνιάω, καταφρονέω, κατεξουσιάζω, κατισχύω, alike differ 
from their simplicia in that they indicate action unfavourable to an object. 
(The bringing in of an object, where the simplex had been intransitive, 
is found not infrequently, but it is not easy to connect it causally with 
the preverb: further instances are katapyéw put out of action, from ἀργέω 
be out of action, καταπονέω, κατασοφίζομαι, κατακληρονομέω, κατακαυ- 
Χάομαι, καταγελάω.) In a good many of the verbs classed as perfective 
we could easily reach that sense through xara=down, and in others 
by the “ hostility’? connotation: it is less easy to find compounds 
where we might recognise the meaning which Brugmann regards as 
most original. Κατακολουθέω and καταδιώκω describe following right 
over an intervening space till the quarry is reached: καταντάω and 
perhaps κατευθύνω are not very different, nor is the ἅνλ. κατεπέστην 
(“‘ make a dead set upon”’). 


1Thumb Hdb. p. 106. 


§ 121-122] WORD-COMPOSITION—PREPOSITIONS. 317 


(b) About 40 compound nouns or adjectives found in NT attach 
themselves to the verbs under (a), as derivatives or as parallel forma- 
tions. On ἀκατάπαστος as a negatived verbal from καταπαύω see ὃ 37. 
Καθηγητής from καθηγέομαι deduco goes into the perfective list, as does 
καταπέτασμα from καταπετάννυμι, unless that is to be compared with 
καταχέω pour over. Κατόρθωμα is from κατορθόω, where κατά has pro- 
duced a transitive verb. Κατατομή mutilation is linked with perfective 
κατατέμνω (we say “cut wp”). Κατάρα against the more general dpa 
has the κατά of hostility, and so κατάθεμα, κατήγορος (-wp, “ρέω, “ρία) 
and καταδίκη. Κατήφεια, of quite uncertain etymology, seems to 
contain xarad=down. Κατάδηλος and κατάλοιπος have the intensive 
κατά. Κατείδωλος is (naturally enough) 4.A.: the special Jewish use 
of εἴδωλον is sufficient to explain this. But it is coined (by Luke or some 
Jewish predecessor) in accordance with analogy: cf. κάθαλος full of salt, 
κάτοξος drenched with vinegar (both New Comedy), κατάγλωττος talkative 
(Epictetus), carddevdpos densely wooded (Nymphodorus—of an uncertain 
date B.C.). 

(c) Kataxpynpvilw=cast κατὰ κρημνοῦ, καθημερινός from καθ᾽ ἡμέραν 
daily, καταχθόνιος -εκατὰ χθονὸς dv, belong to Cl. 11. Compound 
prepositions or adverbs include κάτω (with κατώτερος), ὑποκάτω, κατε- 
νώπιον, καθεξῆς. Adverbial phrases which are always or often written 
as single words are καθ᾽ 6, καθ᾽ ἅ, καθ᾽ ἅπερ, καθ᾽ ὅτι, κατὰ μόνας, καθ᾽ 
ὅλου : it makes no real difference whether we make them one word or 
two. Καθώς--καθ᾽ ὥς belongs to the same category as ἐκ τότε etc. 
(Brugmann Gr.* 524). 


§ 122. 11. *t Μετά has been already noted 
as an extension of μέ with the same element 
that we find in κα-τά. The I.E. me seems to have produced 
medhi, whence medhios=pécos, medius, mid, Skt. madhyas, etc. : 
the Germanic preposition seen in Goth. mip, Ger. mit, might 
equally go back to *meét. The Greek form with -ta, accented 
μέτα as an independent word, has apparently no parallel, 
since the Iranian mat “with” must be compared with ἅμα 
because of its obvious link with Skt. smat: see Brugmann 
Grd.2 τι. ii. 856. Μέςχρι(ς) and the dialectic μέστα, μέσποδι, 
μέττ᾽ és, μέσφα, with the early compound μεταξύ, belong to the 
family. It is possible to conjecture that mett was the oldest 
form, accounting for Germanic and Greek alike: in that case 
*weés (cf. the dialect forms above) may be its surviving repre- 


Meta 


1 Καθολικός, from καθ᾽ ὅλου, does not occur in NT, though appearing in 
late MSS. 


318 A GRAMMAR OF NEW TESTAMENT GREEK. [§ 122-123 


sentative, like πρός from προτί, and μέτα may be due to the 
analogy of κάτα. As to its meaning, it seems to start from 
amid, as we may see iN μετέχω “ possess among others, share.” 
The common sense of change is assigned by Brugmann-Thumb 
Gr. 509 to the local idea of an intervening space (medium) which 
is crossed : they compare Lat. interesse “ differ.” As a preposi- 
tion the primary force is with, m various senses, which alone 
survives in MGr.,1 except for phrases of manner: after, coming 
from the idea of crossing over the “ mid” space to the other 
side, is of course common in Hellenistic, but leaves no trace 


to-day. 


(a) Mera forms 21 verb compounds in NT. Of these 16 have the 
idea of change, answering often to Latin compounds with frans-, which 
supports the explanation of this force of μετά given above from Brug- 
mann. I include μεταμέλομαι here, though Grimm finds the sense of 
“* afterthought”: there seems no reason for placing it and μετανοέω in 
different classes, and the latter indicates “change of mind” beyond 
question. “Sharing” action (like German cpds. in mit-) accounts for 
μεταδίδωμι, μεταλαμβάνω, μετέχω. Μετακαλοῦμαι and μεταπέμπομαι, 
** call for * and “ send for,” have the sense which appears with μετά c. acc., 
action crossing over a space to a goal. 

(65) Some 16 noun compounds of μετά are found, if we include 
μετέωρος on account of its derivative perewpiLouar. Of these 11 are 
derivatives of verbs included under (a), while μετοικεσία may be added 
as very near to peroxile. Μεθόριον (Mk 77 ANw) is a literary word 
(μετά between) foisted on Mark by copyists. Μεθοδεία comes from 
μεθοδεύω, and this from μέθοδος, where pera is like that in μεθέπω “ follow 
afier, up.” Μέτωπον is apparently the part “ between the eyes” (ef. 
μεσόφρυον). Μετέωρος (cf. Lesbian πεδ-ἄορος --- from ἀξείρω) is like 
μετάρσιος (from aipe): in μεταίρω the idea of removal is clear, so that 
“* lifted across” a gulf of air will be the starting-point. 

(c) Μετέπειτα afterwards is a very early example of pera=after, 
occurring as it does in Homer. Cf. the note above, p. 293, on this early 
(non-Attic) justification of Hellenistic combinations like ἀπὸ τότε, ete. 


Feed § 123. 12. ἘΠῚ παρά makes itself conspic- 

= uous by retaining prepositional use with all 

three cases : it has, however, a very limited range in MGr. It 

belongs to a family with the common element per, including 

περί (loc.), πάρος (gen. abl.), wapai—Lat. prae (dat.), πέραν, 

πρό, πρω-, πρός (Ξ-προτῶ : our own for, before, from and fro, 
τ Thumb Hdb. 104. 


ow Δ᾽ 


§ 123) WORD-COMPOSITION—PREPOSITIONS. 319 


Latin pro, prae, per, por-, may be mentioned among words 
quotable from all I.E. dialects. The root of what may well 
have been a primitive noun can probably be recognised in 
πόρος, our fare, πείρω “ pierce”: “ going across” (cf. πέραν) 
is the central idea. From the adj. *peros we have in Greek the 
acc. fem. πέραν to the other side (c. gen.), πέρω instr. beyond : 
see Brugmann Gird.? τι. 11. 8721. Πάρα itseli—whose inde- 
pendent accent is unoriginal, like that of «ara—is presum- 
ably an instr. case-form, but such identifications of prehistoric 
words are of course only conjectural. It answers to Skt. pura, 
Av. para “before,” Alban. para, Olr. ar, Goth. fatra, all= 
“before,” in time or place. Gothic faviv may answer in form to 
the alternative πάρ, Lat. por-: (παράγων) παρὰ τὴν θάλασσαν 
is in Wulfila favir marein, (ἔπεσε) παρὰ τὴν ὁδόν fair wig. 
“Close to, beside” is apparently the oldest Greek meaning ; 
but new developments branch out: beside gives mis-, stealthily 
(gs. entermg by a side way), and sometimes diverges into 
onward, with metaphorical application. 

(a) Tapa forms 53 or 54 verbal cpds. in NT, a fair proportion of 
which are not very easy to define in relation to the preverb’s earliest 
meanings. The simplest are those where we may represent the παρά by 
‘* beside, close to.” Thus παρακαθεσθεῖσα in Lk 1039 is like 947 ἔστησεν αὐτὸ 
map éavta. Παραβάλλειν compare (Mk 4°° ADw—a plausible reading), 
παράκειμαι, παρακολουθέω follow closely, Tapapevw, παρατηρέω watch 
closely, παρατίθημι, παραχειμάζω, παρίστημι, παροικέω; are fairly clear. 
Others have the idea of motion affecting the sense, so that we compare 
the uses of wapac. acc. Παραβιάξομαι (in 1λιΚ6) -- βιάζομαι παρ᾽ ἐμαυτόν, 
Tapayivouat—=come near, Tapept=am near, Tapatuyxdvw=chance (to 
be) near; παρακαλέω (in Ac 287°) call to one; παραδίδωμι hand on to; 
παρακύπτω stoop close to; παρασκευάζω prepare ready—almost identical 
with σκευάζω but suggesting a “ presentation”’ of the prepared object. 
The idea of going past does not come naturally out of the general sense 
of closeness, and may go back to earlier elements (cf. praeter). So παρα- 
βαίνω, παρέρχομαι, παράγω, παραπλέω, παραπορεύομαι, παραρρέω, 
παραφέρω, παροίχομαι. Aside, developing into mis-, is recognised in 
παραθεωρέω overlook, παραιτέομαι deprecor (qs. “ask aside’’), παρακούω 
mis-hear, ignore (also with παρά close=overhear), παραπίπτω fall aside, 
παραλογίζομαι mis-calculate. Others have the ablatival connotation : 
Tapadexopat, παραιτέομαι (in Mk 15°), and παραλαμβάνω suggest the 
source (παρά τινος) as παραδίδωμι does the recipient (παρά τινι). A kind 
of dative idea attaches to παραινέω and παρακαλέω, where παρά suggests 
an intimacy with the object. Παραγγέλλω is pass a message on, with 


320 A GRAMMAR OF NEW TESTAMENT GREEK. [§ 123 


the same idea of “‘ onward motion” that we saw in παραδίδωμι, etc. “On 
one side” may possibly be the meaning in παρακαλύπτω and παραλύομαι. 
The “ἡ onward” nuance seems to produce παραξζηλόω, παραπικραίνω, 
παροξύνω, mapopyitw and παροτρύνω. Παρακαλέω and παραμυθέομαι 
with the sense “ἡ comfort”? may be attached better to the idea of in- 
timacy, speaking “to the heart.” “Sideways” and so “ stealthily ” 
is seen in the παρ-εις- epds. of ἄγω, δύω and ἔρχομαι: the last, with 
παρεισφέρω, has also the meaning besides. Παρεμβάλλω as a military 
word, said to be Macedonian—i.e. coined especially in Philip’s or Alex- 
ander’s army—amust be left uncertain, as a t.t. which might have origin- 
ated in conditions we cannot trace. On παραφρονέω, παραδειγματίζω and 
παρομοιάζω, see (b). 

(b) Twenty-six noun compounds of παρά in NT are formed from verbs 
already occurring there. Among these we should notice παρουσία, in 
which the RV marginal note (“ Gr. presence’’) would suggest that the 
idea of “ἡ motion towards” is to be excluded: outside evidence for the 
technical meaning “ royal visit ” shows that advent is as literal a rendering 
as presence, which occurs in some places. The simple locative force may 
be seen in eUmdpedpos (cf. παρακαθέζομαι), παράδειγμα (from παραδείκνυμι 
show two things side by side) whence wapadetypatilw, παραλλαγή (devia- 
tion), παράσημος (marked on the side), παρεπίδημος (staying with), 
παρηγορία (cf. note on παραμυθέομαι, “tov above), πάροδος (journey past), 
παροιμία (cf. obiter dictum, τὸ παρ᾽ οἶμον), πάροινος (-- παρ᾽ οἴνῳ ὦν), 
παρόμοιος (strengthened ὅμοιος, παρά as elsewhere =beside in comparison) 
whence παρομοιάζω,3 παροψίς (side-dainty—with transference to the dish, 
the converse change to that in our dish when used of the food). Mapa- 
βολεύομαι, from παράβολος venturesome (cf. class. παραβάλλομαι in same 
sense), has the verbal part expressing the energy of βάλλειν, instead 
of being static as in παραβολή. Hence παρά describes motion along, like 
that of an object flung into a rushing stream. Παραφρονία from 
mapappev=having the mind awry: tapadpovéw might be a cpd. of 
φρονέω, ** to think awry,” but is more probably from παράφρων. 

(c) Παράδοξος --παρὰ δόξαν ὦν, and παράνομος (whence -ία) -- παρὰ 
νόμον ὦν, are obvious Class VII. forms, with παρά (c. 860.) -εσοπέγατγῳ to. 
Παραθαλάσσιος -- παρὰ θαλάσσῃ ὦν, and παράλιος -- παρὰ ἁλὶ Sv, show the 
ordinary locative sense of παρά. Παραχρῆμα is a phrase adapted un- 
changed for adverbial use. In παραπλήσιον, παραπλησίως and παραυτίκα 
the addition close by intensifies the force of an adverb which invites an 
element of this kind. 


1 The simplex δειγματίζω is apparently later than the epd. in emerging, and 
is best taken as a popular derivative from παραδειγματίζω on the basis of the 
existing δεῖγμα. But that the verb does not emerge till the papyri is not 
positive proof that it did not exist. [See also Vocab. 138.—Ep.] 

* The simplex, if it really exists, is secondary to the epd. 


§ 124] WORD-COMPOSITION—PREPOSITIONS. 321 


§ 124. 13. *t Περί (πέρι) answers exactly 
in form and meaning to Skt. pdri, Av. pairi, 
“around”; Lat. per, Goth. fatr-, Lith. pe, less obviously 
connected in meaning, remind us that the semasiology will 
not be wholly easy to follow. Brugmann (Grd.? τι. 11. 865) 
notes that the divergence dates from I.E. times. He gives the 
meanings of the preverb as hiniiber, tiber etwas hinaus— 
Ubergang in ein anderes Verhdltniss, in eine andre Form, zu 
Entgegengesetztem, Uménderung, besonders zum Schlechten— 
hindurch—umher, herum, um—durch und durch, allseitig, sehr. 
But in Greek, as he and Thumb observe (Gr.4 511 f.), around 
covers most of the ground : its development was connected with 
that of ἀμφί, and in dialects and Κοινή the preposition came 
near ὑπέρ. 


Περί 


(a) There are 32 περί verbs in NT (33 if we accept περιραίνω from 
x* in Rev 1915), of which about four-fifths show the meaning “‘ round, 
about.” Περιάπτω kindle is not quite clear: perhaps it is enough to 
compare περιαστράπτω and περιλάμπω and point to the way the flame 
runs “ἡ round.” Περιβάλλω clothe illustrates the link with ἀμφί (cf. 
ἀμφιέννυμι) : so περιζώννυμι (=Lith. pérjosti). Περιρήγνυμι tear off 
(clothes) is correlative. Περιπατέω walk about links itself with Lat. 
perambulare. So does περιφέρω bear about with perfero. To other 
headings we assign the following. Περιαιρέω remove is taken by Grimm 
to include (2 Cor 3, Ac 274°) the idea of something enveloping or lying 
on both sides ; but itis better to make it intensive (see above). Περιεργά- 
Copa (cf. περίεργος) implies overdoing a thing, with the pejorative nuance 
included under Brugmann’s second heading (above): cf. perverto, perdo, 
perimo. Περιέχω (in 1 Pet 28, and vernacular sources) has a curious 
intransitive force, is contained: the active include is easy, and the in- 
transitive use appears to follow the analogy of otras ἔχει etc. Cf. 
περιοχή period. Περιίσταμαι avoid, a Kown use, is somewhat difficult: 
it may belong to the second category again, of “‘ passage into another 
relation,’ whence shift round, mid. intransitively with acc. of object 
shunned. Περιλείπομαι remain over, survive, περιποιέομαι make survive 
for oneself, gain, have the force seen in περισσός. Περιμένω is like 
pervenio, Goth. fairrinnan (-- ἐφικνεῖσθαι) : περί is hindurch. Περιπείρω 
pierce through, like perforo etc. Περιπίπτω is apparently fall amid, a 
kind of passive to περιβάλλω. Περισπάομαι distract is like our “ pull 
about.” Περιφρονέω despise is compared by Brugmann with Skt. paricaks 
(caks =see—cf. περιορᾶν) overlook, disregard. 

(Ὁ) Ten noun compounds in NT are linked with verbs included under 
(a). This includes edmepiotatos, the difficult ἁ.λ. discussed above, ὃ 106. 


322 A GRAMMAR OF NEW TESTAMENT GREEK. [ 124-125 
Περικάθαρμα and περίψημα (* Wyo, Wao, wipe), start from “ cleaning, 
wiping round.” Mepixpatys=having κράτος over, lit. around. Περίλυπος 
is intensive, having excessive sorrow. Περίοικος having house around—like 
περίχωρος. ἰ5 really independent of περιοικέω, but is counted among 
the ten abcve. Περιούσιος 1 is said to be a Biblical coinage (LXX), like 
ἐπιούσιος in NT, to render a Hebrew word meaning “ special possession.”’ 
Can we get the meaning special, superior, out of περιουσία superiority ? 
It seems desirable to connect it with the quite common noun if at all 
possible. 

(c) Περικεφαλαία is a Cl. VII. noun=70 περὶ κεφαλήν (cf. προσ- 
κεφάλαιον). Πέριξ appears to be the nom. sing. of an old adj. formed 
directly from περί with suffixal x: so περισσός (=perikyos) and its 
derivatives. 


8 125. 14. Ἔ Πρό, for the cognates of which 
see on παρά ad. init., answers to Skt. pra, 
Ay. fra, Lat. pré-, ΟἿ. ro-, Goth. fra-, Ger. ver-, E. for-, Lith. 
pra: its original and still normal sense is forwards, before, of 
either time or place. Lat. prd, prad-, is closely related, and 
in many cpds. takes the place of pré-. The I.E. preverb seems 
to have been used with special frequency: its close attach- 
ment to the verb is marked by the rule that in double or treble 
epds. containing pré this always comes last—e.g. ὑπεκπροφυγών 
(Hom.), Skt. dnu pra eti, ete. (This rule is primitive, but 
extinct fairly early : five NT epds. violate it.) Significant also 
is the frequency of epds. which seem to have I.E. antiquity, 
such as pré gem (Gk., Lat., Goth.), pré bher (Ar., Gk., Lat., 
Goth., Slav.). 


(a) Πρό forms in NT 49 verbal epds., in 28 of which it more or less 
clearly implies before, of time, and in about 11 before, of place: in 
προγράφω, προέρχομαι, and προοράω both occur. Besides these we have 
verbs in which forward or forth would suit better, whether in local sense, 
aS προπέμπω, προτείνω, προφέρω, or metaphorically, as προκαλέομαι, 
προλέγω (in some disputed places), mpotiBepat—propono, προτρέπομαι. 
In προαιρέομαι we have the idea of preference (cf. πρὸ πάντων), in προ- 
totapat? that of protection or care. (Προτάσσω is counted in the list on 
the strength of one occurrence in D*.) The problem of προεχόμεθα in 
Rom 3° does not turn on the προ- : see the Verb-syntax. 

(b) Five noun compounds are linked with NT verbs accounted for 


, 


Πρὸ 


1See J. B. Lightfoot, Fresh Revision (1891), pp. 260 ff. 
2 Brugmann-Thumb (p. 514) prefers forward, 1.6. openly; but does this 
suit the following genitive so well ? 


Olive ox. 


§ 125-126] WORD-COMPOSITION—PREPOSITIONS. 323 


under (a). Before in time or space appears in πρόγονος, πρόδρομος, 
προθεσμία, πρόκριμα, προπάτωρ ; openly in πρόδηλος, forward in πρόβατον 
(whatever the second element), πρόθυμος (=having mind forward, ready), 
προπετής, πρηνής (if headlong 1). Προδότης traitor, from mpodidope= 
prodo, seems to start from forth, away: we use give away in the sense of 
betray. Πρόφασις is saying or showing in front of, i.e. an excuse to cloak 
the reason. Προφήτης has representative force, an interpreter speaking 
for the divinity. According to Brugmann (Gr.* 158) we should place here 
the problem-word πρότσφατος, the second element of which is an isolated 
by-form from the root of φθάνω : this is obviously commended by the 
meaning. 

(c) Πρό has a comparative πρότερος : πρῶτος if for πρώξ-ατος is from 
a cognate and not directly formed. To Cl. VII. belong προαύλιον -- 
τὸ πρὸ αὐλῆς ov, and πρόχειρος (whence “ρίζομαι) -- πρὸ χειρῶν ὧν ΟΥ̓ ἔχων, 
with πρό local in both cases. Προσάββατον from πρὸ σαββάτου has πρό 
temporal. Πόρρω may be remotely connected. 


§ 126. 15. ἘΤΤ Πρός is less conspicuous as 
a preverb than as a preposition. Its oldest 
Greek form is mpoti=Skt. prdti, but the dialects show also a 
synonymous ποτί (Doric) and πός (Arkad.-Cyprian). The Ar. 
pas-ca (“after”), Lat. pos-t, Lith. pas (“near, to”), Alb. 
pas (‘after’), may prove pos as well as poti to be pre-Greek, 
in which case we have an additional force to explain the 
survival of πρός (=proty, the sandhi-form before vowels) over 
προτί. Like Greek, Aryan had both prati and pati, the former 
surviving in Indian, the latter im Iranian: whether there was 
any distinction of meaning we cannot tell. Both seem to 
have been local adverbs=“ over against.” IIos survived in 
the Kowy of Phrygia, but otherwise only πρός is found in 
our period. For all these statements see Brugmann-Thumb 
Gr. 514. Outside Aryan there is hardly any sign of προτί: 
Brugmann (Grd. τι. i. 877) would find its gradation doublet 
preti in Lat. pretium, and Lettish pret (“ against”), also Pam- 
phylian περτί. He suggests that po-s and po-tv may be ex- 
tensions of (α)ρό, as pro-ti of pro: for the element -ti cf. ἀντί, 
me-ti (?—see under peta), ἔτι. Dvrection and addition are the 
general headings under which the Greek meanings fall. 


, 


Πρὸς 


1 Which is improbable for the one NT occurrence, since ΒΡ. Chase’s argu- 
ment for πρησθείς swollen as the meaning in Ac 118: see J7'S xii. 278 and 
Harnack’s endorsement in ThLZ xxxvii. 235. 


324 A GRAMMAR OF NEW TESTAMENT GREEK. [8 126-127 


(a) There are 44 composita with πρός in NT, if we include προσαχέω 
(Ac 2727 B*) and προσεγγίζω (Mk 24 ACDw). In all but about 8 of these 
the πρός is directive, answering to the meaning of πρός as preposition 
6. acc. In προσαιτέω and προσδέομαι we should express the force better 
by πρός c. abl., where the case implies from and πρός adds “ to oneself.” 
Sometimes the prepositional equivalent would be rather zpos c. loc., as 
προσμένειν τινί-- μένειν πρός τινι. Other cpds. have πρός -εἴη, addition, a 
meaning which it held in its use as an adverb without a case: so 
προσαναπληρόω, προσαπειλέομαι, προσδαπανάω, προσεάω, TpocepydLopar, 
προσοφείλω, προστίθημι. One of the directive cpds. deserves a note, 
προσεύχομαι, an old word which is markedly appropriated in NT for 
Christian prayer: contrast εὔχομαι in Ac 277°. The vivid sense of πρὸς 
τὸν θεόν accompanying it made it the natural word. 

(b) Sixteen noun cpds. in NT are linked with verbs accounted for 
under (a). ᾿Απρόσιτος from πρόσειμι go to, πρόσχυσις from προσχέω 
pour on, may be added. A pure noun epd. is προσφιλής dear to, pleasing. 
Προσευχή was probably secondary to the verb προσεύχομαι, with εὐχή as 
a starting-point. Πρόσωπον (with its derivatives) goes with μέτωπον, 
and on its analogy will mean the part “‘ up to the eyes.” 

(c) Nouns of Cl. VII. are πρόσκαιρος from πρὸς καιρόν, προσκεφάλαιον 
from πρὸς κεφαλήν, πρόσπεινος from πρὸς πεῖναν. Ἔμπροσθεν from ἐν 
and πρόσθεν : see under Improper Prepositions, ὃ 130. 

§ 127. 16. ἡ Σύν, which as a preverb has a 
most extended use in Greek, has no clear 
cognates outside : μεταξύ is witness to two older by-forms, ξύν 
and €v. It may be assumed that ξύν (cf. Ion. ἕξυνός--- κοινός 
in formation and meaning, being ξύν and κομ- with adj. suff. 
-yos) is the older form, but the phonetic conditions which 
caused the « to vanish are not cleared up: ξύλον and ξύλινος 
have initial σ- in a few Attic inserr. of iv/B.c.1 For the possibility 
of a Slavic cognate see Boisacq s.v. ξύν. It is conceivably a 
primitive Greek compound of ἐξ (reduced by gradation as in 
Lat. s-uper and s-ub) and sw which may be recognisable in 
Lith. δὼ “ with,” unless this is capable of representing ksu by 
itself: see Brugmann Gird.2 τι. ii. 897. The restrictions of 
σύν when used as a preposition are obvious—note that they 
are not shared when the word becomes a preverb. 


Σύν 


(a) The proper meaning οἵ σύν being together, it is capable (like the 
synonymous Latin com-) of making pure perfectives, under which heading 
we may class about 24 of the 123 σύν compounds in NT.?_ The rest con- 


1 Meisterhans? 92. 
* The total includes συνελαύνω (Ac! AE Pw) and συγκατανεύω (Ac! D). 


et 


§ 127] WORD-COMPOSITION—PREPOSITIONS. 325 


tain σύν =with, or various shades of meaning included under together. 
The perfectives reckoned in the above total are συλλαμβάνω (concipio), 
συλλογίζομαι (colligo), συναρπάξω (comprimo), συνευδοκέω (in 1 Cor 7t- 
—elsewhere joint approval), συνέχω, συνθλάω, συνθλίβω, συνθρύπτω (cf. 
confringo etc.), συγκαλύπτω (conceal), συγκάμπτω, συγκινέω (commoveo), 
συγκλείω (concludo), συγκύπτω, συνοράω, συμπεριλαμβάνω, συμπίπτω 
(concido), συμπληρόω (compleo), συμπνίγω, συντελέω, συντέμνω (concido), 
συντηρέω, συντρίβω, συγχέω (-χύννω) (confundo), συσπαράσσω (convello). 
Some of these have the ordinary force of σύν in some of their uses; and in 
other cases the line between perfective σύν and non-perfective is not easy 
to draw. A few of these latter call for note. Συμβαίνω (cf. contingo) 
seems to start its special sense from the idea of coincidence, and συμφέρω 
from contributory action. Συναλλάσσω when compared with the nearly 
synonymous καταλλάσσω might be treated as an instance of σύν ‘per- 
fective ; but it seems to derive its force mainly from the idea of inter- 
course. Συνελαύνω, in the one place where NT MSS show it (Ac 7°, 
where the tense alone suffices to make the reading impossible), must 
be compello=drive together, of two parties. Συνίημι and συνιστάνω 
(exhibit) seem to contain the idea of a mental “‘ synthesis.” Συλλυ- 
πέομαι in Mk 3° is difficult, since the word in its earlier record involves 
sympathy, which is somewhat forced here: can it be perfective =wtterly 
distressed ἢ Συγχράομαι might almost be called an equivalent of 
xpacba=deal with; but the sociative σύν seems to be consciously 
present, and there is no specially perfective force visible. 


(Ὁ) In noun epds. σύν acts very much as in verbal. There are 18 
words and groups in NT closely connected with verbs found there: 
σύγχυσις, συνοχή, συντέλεια, συντόμως, σύντριμμα belong to the per- 
fectives, and the rest to the com- class. (Συντέλεια of course is from 
συντελής, but its Hellenistic meaning, conswmmation, appears to spring 
from association with συντελέω.) Definitely verbal cpds. are συναγωγή 
(which has become concrete from abstract) and its cpd. ἐπισυναγωγή, 
συγγνώμη (from συγγιγνώσκω pardon)—the special meaning seems to 
spring from the idea of sympathy, συγκυρία (from συγκυρέω coincide), 
συνδρομή (cf. συντρέχω), συνείδησις (a Hellenistic derivative from 
σύνοιδα --οοηδοῖο), σύνεσις, συνετός and ἀσύνετος (from συνίημι), 
συνωμοσία (from συνόμνυμι --οοπῖατο), συστατικός (from συνιστάνω -- 
commendo), and ἀσύνθετος (from συντίθημι). Of purely noun character 
are a good many of Cl. V. (based on IV.c), as συγγενής having common 
race, σύμμορφος having same form, σύμφωνος (whence °€w etc.), σύμψυχος, 
συμπαθής (whence °€w), σύζυγος, σύμβουλος, (συνέδριον from) σύνεδρος, 
συνεργός, (συνήθεια from) συνηθής, σύντροφος (τροφή), σύνδεσμος, etc. 
Others again are Descriptives (IV.c), as συγκληρονόμος, συγκοινωνός, 
συμμαθητής, συμμιμητής, (συμπόσιον from) συμποτής, συναιχμάλωτος, 
σύνδουλος, συνέκδημος, συστασιώτης. Note the pleonasm in συμμέτοχος. 


326 AGRAMMAR OF NEW TESTAMENT GREEK. [§ 127-128 


(c) Apart from μεταξύ (see above), there seem to be no compound 
adverbs containing σύν ; nor are there any Cl. VII. formations in NT. 


§ 128. 17. *t Ὑπέρ was in proethnic speech 
the antithesis of ndheri (under, Lat. infer-us, 
infra, Skt. ddharas, Av. adairi, perh. ἀθερίζω “to slight ”’). 
The alternative form with the locative suffix -ὖ appears in 
Skt. wpdri (also Iranian). Lat. s-wper (a compound), Gaulish 
Ver(cingetorix etc.), Goth. ufar=over, and Armen. ὁ ver, 
illustrate it further. The meaning is uniform throughout the 
wide field, viz. over, passing into across, beyond. Ὕπερ is still 
capable of adverbial use in the NT period (2 Cor 1133). For 
the curious fact that the ILE. wper(?) and το, which are 
obviously kin, should in Greek and other languages have anti- 
thetic meanings, see under ὑπό. 


Ὑπέρ 


(a) Only 14 verb epds. of ὑπέρ occur in NT; but it is conspicuously 
capable of forming new ones in this period. Thus it is very likely that 
Paul coined ὑπερνικάω. In this word and in ὑπεραυξάνω, ὑπερπερισσεύω, 
ὑπερπλεονάζω, ὑπερυψόω (all Pauline) the preverb simply magnifies, as 
in overjoyed (ὑπερχαίρω). Elsewhere, as usually in our own over- cpds., 
there is the sense of excess: so ὑπεραίρομαι, ὕπερβάλλω (cf. overshoot), 
ὑπερεκτείνω, ὑπερεκχύννομαι, ὑπερφρονέω (cf. overweening). Ὑπερβαίνω = 
go beyond, transgress; Smepopdw—=overlook, neglect; ὑπερεντυγχάνω Ξε 
intercede for (ὑπέρ τινος); bmep€xw=surpass. It is characteristic of 
Paul’s temperament that only 3% occurrences of ὑπερ- cpds. in NT lie 
outside the corpus Paulinum. 


(ὁ) Ὑπερβολή and ὑπεροχή are parallel in meaning with their verbs 
in (a) above. “Yrepypavos—“< overbearing”? is generally assumed to 
contain ὑπέρ (or depos) and the root of daive, but the -7- (for a, as 
Pindar shows) is not thus satisfactorily explained.t Nor is ὕπερῷον, 
from the adj. ὑπερώιος upper, the suffix of whichis obscure. Ὑπέρογκος 
with excessive swelling is a Descriptive based on a Cl. 1V.c cpd. 

(c) Ὑπέρακμος, from ὑπὲρ ἀκμήν beyond prime, belongs to Cl. VII. 
Ὑπερέκεινα, which is guaranteed to be good Kow7 by the strictures of 
Thomas Magister, is simply ὑπὲρ ἐκεῖνα turned into one word: so the 
earlier ἐπέκεινα. In ὑπερεκπερισσοῦ a compound adverbial phrase has 
been heightened as in ὑπερεκπερισσῶς, ὑπερλίαν and ὑπερπερισσῶς a 
simple or compound adverb. Ὑπεράνω is rather different, as the former 
element prevails and makes the whole an improper preposition differing 
little from ὑπέρ c. gen.: cf. ἐπάνω. 


1 Wackernagel Dehnungsgesetz 42 makes it=vtrep-i¢-a-vos, with the vowel 
of ν᾽ αφ lengthened at the juncture: cf. κατηφής and derivatives. 


ar ieee 


§ 129] WORD-COMPOSITION—PREPOSITIONS, 327 


§ 129. 18. *f‘yné=Lat. sub, sup- (like 
super, somewhat doubtfully explained as for 
z-upo, x-uper, from ex), Skt. and Av. wpa, Gaulish Vo(retus 
etc.), Olr. fo, Goth. uf, Ger. auf. The adv. ὕψι (super. 
ὕψιστος) and noun ὕψος seem to have affected the originally 
distinct ὑψηλός, for *bEnAds (cf. Gaul. Uxellodunum ‘ Hightown’). 
These cognates, together with ὑπέρ, need to be brought in 
here to account for the meaning: add wp=O.E. uppe, Olcel. 
upp, probably from proto-Germanic wpn, Skt. wpan-ayati 
(Brugmann Grd.? τι. 11. 911). The original idea accordingly 
seems to be “upwards from below”: cf. also Goth. cup 
(I.E. eupn-) “ upwards,” Lat. swsque deque “up and down,” 
sustineo “hold up.” Brugmann also compares aufsteigen, 
succingere, for the idea of motion from below. In that case it 
was possible, in languages which like Greek and Latin had lost 
the I.E. antithesis of over and under (uperi, pdheri—see above 
by the disappearance of the latter, that the starting-point of 
upo should be isolated, and ὑπό sub become the antithesis of 
ὑπέρ super. This is essentially Brugmann’s explanation (op. 
cit. 912). Note that Goth. uf renders ὑπό c. acc., as in Mt 88, 


« 
Ὑπό 


(a) The 27 ὑπό epds. occurring in NT can mostly be paralleled with 
Latin words containing sub. The idea of ‘‘ submission” accounts for 
ὑπακούω,: ὑπείκω, ὑπέχω, ὑποτάσσω ; that of ‘‘ underhand ” for ὑποβάλλω, 
ὑποκρίνομαι.: Under in the literal sense gives ὑποδέομαι and ὕποστρωννύω 
(cf. Aryan, Lat., Goth., Olr., for an identic cpd.). Motion ὑπό τι may 
explain ὑποδέχομαι, ὑποστρέφω, and perhaps originally ὑπάγω, ὑποχωρέω, 
but ὑπό leaves no sensible force but away, or even back, in the last three 
named. ‘“Yravtdw—=come up to; ὑπάρχω spring wp, arise ; ὑποδείκνυμι, 
ὑπομιμνήσκω and ὑπονοέω (cf. suggest) convey the idea of thoughts making 
their way wp into the mind ; ὑποζώννυμι gird up, cf. succingo, ὑπολαμβάνω 
=take up, in various senses; ὑπολείπω (-AtuTdvw)—=leave behind (qs. 
at the bottom); Gmopzévw=undergo, sometimes=tarry behind, as in 
ὑπολείπω ; ὑποπλέω --581] wnder, close up to, and so ὑποτρέχω ; ὑποπνέω 
=sufflo, breathe softly ; ὑποστέλλω =draw back (cf. on ὑποχωρέω above)— 
note that Wulfila renders ufslawp in Gal 215, 

(Ὁ) Ὑπακοή and ὑπήκοος, ὑπάντησις, ὕπαρξις, ὑπόδειγμα, ὑπόδημα, 
ὑπόκρισις and -κριτής, ὑπόλειμμα, ὑπόμνησις, ὑπομονή, ὑπόνοια, ὕπο- 


1 Τὴ both of these verbs the sense of ὑπό has continued to work in the later 
development. ‘Yzakovw was originally only to “answer’’ the door, ὑποκρίνομαι 
(in Attic—developed from “‘ answer ”’) to “ play a part”’: in both we recognise 
originally the local sense found in ὑποχωρέω. 


328 A GRAMMAR OF NEW TESTAMENT GREEK. [§ 129-130 


στολή, ὑποταγή and ἀνυπότακτος are accounted for above, under their 
allied verbs in (a). Ὑπηρέτης (whence -€w) was originally a t.t. of the 
galley (cf. ἐρέσσω), as ὑπηρεσία still shows in classical writers, but it 
early lost its special connotation and became an “‘ wnderling” in general. 
Ὑπολαμπάς (Ac 208 D)=window is a very rare word (see LS, and add 
Syll. 588" (Syll.2 om.) [ii/B.c.]): apparently a screen under which 
the light shines. Ὑπόστασις in its various meanings runs parallel with 
its Latin equivalent substantia, an underlying foundation being implied. 
Ὑποτύπωσις is like Swoypappds, both suggesting a copy traced over 
a pattern below. Ὑπεναντίος is a stronger form of ἐναντίος, with ὑπό 
as in ὑπαντάω : the opponent ‘‘ comes up against” one. 

(c) Nouns of Cl. VII. are ὕπανδρος -εὑπ᾽ ἀνδρὶ οὖσα, ὑπόδικος ΞΞ- 
ὑπὸ δίκη ὦν, ὑποζύγιον -ὑπὸ ζυγῷ ὄν, ὑπολήνιον -εὑπὸ ληνῷ ὄν, ὑποπόδιον 
Ξεὑπὸ ποσὶν ὄν, and ὑπώπιον (whence ὑπωπιάζω) --τὸ ὑπ᾽ ὠπί, the cheek 
just under the eye. All come from the meaning wnder, as used (in class. 
Gk.) with locative. Ὑποκάτω down under is the only compound adverb 
—see the improper prepositions. 


Ill. IMPROPER PREPOSITIONS. 


§ 130. 7 Ἅμα is an old instr. case of sem 
(eis), =smma: Lat. und has a similar history. 
‘Opod is a case of *ouos (=same), which only 
differs in gradation, but it is not used prepositionally. The 
meaning is that of simul (a derivative of the same root), 
with one thing, together. 

*"aveu without is of doubtful history. It has been con- 
nected by Wackernagel with Lat. sine, by others with Goth. 
anu (Ger. ohne). So Brugmann-Thumb, Gr. 523. 

*"avrixpus (Attic—other dialects are without the -s, for 
which see under ἄχρι(ς)) opposite. Clearly a cpd. of ἀντί: its 
second element has been assigned to the root of κάρα, qs. 
“having the head (face) opposite”: this is not perhaps a 
great improvement on the older derivation from xpovw. WH 
accentuate proparoxytone: Blass (p. 20) gives Attic avtuxpos= 
downright, but accentuates this late preposition on analogy 
of Attic (κατὴ)αντικρύ. 

*°Avtiunépa opposite has the later form without final v: it is 
a epd. prep.—see below for its elements. 

* ᾿Απέναντι opposite : see ἔναντι below. 

**atep without probably came into the Kovw7 mainly from 


Improper 
Prepositions. 


§ 130] WORD-COMPOSITION—PREPOSITIONS. 329 


Ionic, which would account for its limited use there [LXX, NT 
and papp. once each]. It is generally taken as=Eng. sunder, 
I.E. sytér. The aspirate destroyed by Ionic psilosis is plausibly 
recognised in ἅτερος, the older form of ἕτερος (cf. Attic θάτερον) 
=separate (“ sundry”). Sine and perhaps ἄνευ will be cognate. 

* "aypr(s) until is a phonetic doublet of μέχρι(ς) : α (=m) is the 
weak grade of we. The final -s, which had become functionless 
in the prehistoric period and came into Hellenistic perhaps 
through the influence of Epic language, appears to a very 
limited extent m the NT as m LXX and papyri, and never 
before a consonant. See Thackeray 136, WH App.? 155: for 
its history Brugmann KVG 456. There is no visible difference 
between ἄχρι and μέχρι. Brugmann gives the Armenian merj 
“near” as a cognate. 

ἜΤ Ἐγγύς near (cpve. ἐγγύτερον) is the neuter of an obsolete 
adjective in -ὖς with an -s added which has the same history as 
that in ἄχρις, οὕτως, εὐθύς etc. No very safe guess has been 
made as to its etymology. 

* "Exrdés outside is a derivative of ἐξ: the -τος is a suffix with 
ablative force as in Lat. coelitus, Skt. namatas, etc. See Brug- 
mann KVG 455, also 180, where Locrian ἐχθός is noted as 
older, ἐξ being for *éys (cf. ἔσχατος). 

* Ἔμπροσθεν in front of is a combination of ἐν with πρόσθεν, 
that is πρὸς + ablative suffix -θεν. 

*"Evavte in the presence of has two further compounds 
ἀπέναντι and κατέναντι, which secured a place in Kow7 litera- 
ture, while ἔναντι itself appears in a very old Cretan inscription, 
and in LXX and one or two inscriptions and papyri of the 
Hellenistic period. Wackernagel (Hellenistica, pp. 3-6) shows 
that it came into Hellenistic from Doric Greek. The Attic was 
*éyavtiov, which held a stronger position in the Κοινή than an 
element drawn from the less influential dialect. It is ace. of 
the adj. ἐναντίος (see p. 308). ᾿Απέναντι has the same three 
adverbs as the French en avant, but in a different order. (Note 
that ἔναντι was for Grimm presumably a Jewish coinage !) 

* “Evexev, less frequently ἕνεκα, also εἵνεκεν (quater) because 
of. On the variant forms cf. Thackeray 82 f., 135, Mayser 241 f., 
Crénert 114, Schwyzer 35f., Nachmanson 18f., WS 50. The 
NT order of frequency matches that found in the Kovvy sources : 
it does not seem possible to trace much system in the use of 


330 A GRAMMAR OF NEW TESTAMENT GREEK. _ [ 130 


variants due originally to dialect mixture. The Attic was 
ἕνεκα, which was invaded by ἕνεκεν as early as iv/B.c. (end) : 
see Meisterhans? 215 f., who shows that in ii/B.c. it has almost 
driven out ἕνεκα even in Attica. Thumb Dial. 359 gives 
εἵνεκα and évexe (Erythrae al.) évexe (Ephesus al.) as the Ionic 
forms. It is derived from *évFexa, and the second element 
is a case-form of 4/ Fex- (whence the ptc. ἑκών). Brugmann 
(Gr.4524) compares Armen. vasn with same meaning as ἕνεκα, 
but notes that its morphology is not clear. 

* ’Evrés within, Latin intus: see ἐκτός above. 

*"Evimov before, in presence of (cpd. κατενώπιον), an adverb 
from the old adjective ἐνώπιος, a derivative of the phrase 
ἐν ὦπα (acc.) to the face, which became stereotyped as an 
adverb. It was only moderately common in the Kown, but 
attained currency among Biblical translators as a conveniently 
literal equivalent of "I : see Prol. 99. In legal papyri it 
gains a separate currency by its equivalence to Lat. coram. 

Ἔξω outside, whence (cpve. ἐξώτερος and) the similarly used 
ἔξωθεν, is best, I think, treated as a compound of ἐξ and the 
obsolete adverb *w, Skt. d, possibly a gradation variant of 
either (Lat.) ὦ or 2: traces of it are probably visible in ὠ-κεανός, 
ὠ-ρύομαι, ὀ-κέλλω, ὀ-δύρομαι al. Cf. ἔσ-ω, ἄντω, κάτ-ω, ὀπίσω. So 
now Brugmann KVG 465. 

* Ἐπάνω above, upon, a compound of ἐπί and ἄνω (see above). 

* "Enéxewa beyond, lit. to that (side), ἐπ᾽ ἐκεῖνα. Cf. ὑπερέκεινα. 

*"Eow (whence ἐσώτερος and ἔσωθεν adv.) inside. The classical 
alternative εἴσω (more correct, since εἰς is phonetically the 
form before vowels) has rather strangely yielded in Hellenistic 
to ἔσω, though ἐς is there obsolete (above, p. 304). For -w see 
above s.v. ἔξω. 

*"Ews until, also of space as far as (as ὡς in MGr, Thumb 
Hellen. 125). It answers (except in the final -s) to Skt. yavat 
“so long as,” with correlative tavat: in Homer these were pre- 
sumably ἄρος and ta@Fos (Doric tas, as), but they are affected 
by Ionic spellings. “Ews is accordingly from the stem of the 
relative ὅς, with a possessive suffix (Brugmann KVG 667): its 
passage from conjunction to preposition (in and after Aristotle) 
is traced by Brugmann (Gr.* 496) to the analogy of μέχρι, 
which already combined these functions. 

* μεταξύ between is most simply taken as a mere compound of 


§ 130) WORD-COMPOSITION—PREPOSITIONS. 331 


μετά and ξύν : cf. νύ by the side of νῦν. (So Brugmann Gr.! 
521). 

* Méxpt(s) until : see ἄχρι(ς) above. 

ἘΌπισθεν and ὀπίσω behind, after. The older form ὄπεθεν 
seems to be an ablative from the noun base whose accus. is seen 
in κατόπιν. But at the same time ἔὄπε belongs with gradation 
to ἔπι (éri)—cf. ὀπ-ώρα, ὀψέ, Latin op (ob) and our after: its 
weak grade is seen in m-é{w. The form of both these “im- 
proper”’ preps. has been affected by the antithetic πρόσω and 
πρόσθεν : πρόσσω (=proty-d, with 6 as in ἔξω above) produced 
its opposite ὀπίσσω as early as Homer, and ὄπιθεν was 
modified in the same period, surviving only in poetry (see 
ἔμπροσθεν above). An alternative account in Brugmann Grd.? 
11. ii. 729 seems less probable. 

+ Παραπλήσιον near to, neuter of an adj. compounded of 
παρά and πλησίος : see πλησίον below. 

* Παρεκτός except : see ἐκτός above. 

* Πέραν beyond is closely connected with πέρᾶ (cf. ἀντιπέρα 
above) : they are case-forms of a pronominal adjective common 
in Skt. (para=other)—cf. πέρυσι (p. 279), perendie, Eng. far. 
See under περί among the Prepositions proper. 

Ἐ Πλήν except is connected by Brugmann (Gr.4 523—see 
KVG 479) with πλησίον : πλήν Twos qs. “ prope ab aliqua re.” 
The final -v, as in other cases (πέραν al.) may be accus. or instru- 
mental ultimately. The root appears to be seq”, as in ἕπομαι 
sequor : socius and secus show the same divergence as πλήν 
and πλησίον. 

* Πλησίον near (Dor. πλᾶτίον), from 4/ pela “ draw near” 
hence πέλας with accent on first syllable of root, and πλᾶτίος 
accented later. The word is of course the adverbial accus. of 
an adj. 

* ‘Ymepdvw above, a compound of ὑπέρ and ἄνω (see above). 

* Ὑπερέκεινα beyond, lit. beyond that (side), ὑπὲρ ἐκεῖνα. Cf. 
ἐπέκεινα, 

Ἐ χάριν for the sake of, on account of, accus. οὗ χάρις, exactly 
like our thanks to, except for the case governed. 

* χωρίς apart from has a suffix parallel with that in ἅλις, 
Megarian ἄνις (=dvev), which may possibly be akin to the 
plural instrumental ending (λόγοις etc.). The root seems to be 
found in χώρα χῶρος “empty space” and (with different 

VOL. II. PART III.—23 


332 A GRAMMAR OF NEW TESTAMENT GREEK. [§ 130-132 


grades) χῆρος “empty of ” and Skt. hetva “without.” The 
meaning apart comes easily from an instr. of such a noun. See 
Solmsen Gr. Wortf. 110 1. 

*Maxpdv is a preposition in P Oxy i. 11318 (ii/a.p.), but in 
NT has ἀπό after it. It is only the acc. sg. fem. of μακρός, 
with presumably ὁδόν understood. Conceivably the conscious- 
ness of words like this may have prompted the new improper 
preposition *s6$év by way of in the “translation Greek” of 
Mt 41° (not LXX).1 


SUFFLXES.? 
WORD-FORMATION BY SUFFIXES. 


§ 131. We must now turn our attention to the second main 
division in the province of word-formation, namely, the study 
of formative elements within words. In the structure of the 
individual word we can generally distinguish between (a) the 
root, (b) the formative suffix (or suffixes), (c) the case suffix in 
the noun, or the personal suffixes in the verb. In addition to 
these there is sometimes (d) a prefix (or prefixes) at the begin- 
ning of the word. (For illustration, reference may be made to 
Giles, 26 ff.) In this section we are concerned primarily with 
the formative suffix, for notes have already been supplied under 
the heading Composition on the origin and meaning of such 
prefixes as we meet with m the vocabulary of the NT. Here, 
again, it is necessary to deal separately with nouns and verbs. 


THE FORMATION OF NOUNS AND ADJECTIVES. 


δ 132. A preliminary distinction must 
be made between root nouns and those nouns 
and adjectives which contain a formative suffix. Root nouns 
are those in which the case suffixes are attached to the bare 
root, that is, to something which is incapable of further analysis. 
They are generally monosyllabic, and are often found as the 


Root Nouns. 


1 [See further, Appendix, p. 459.—Ep. ] 
2 Dr. Moulton’s MS ends with § 130. The chapter on Suffixes has been 
written by the Editor. See Preface, 


§ 132-133] SUFFIXES. 333 


second member of a compound noun. It is usual to classify 
them as they preserve two or more ablaut grades in the case 
forms, or as they show the same grade throughout. 


(a) To the former class belong : 

Nous, gen. ποδός. Dor. πώς. Skt. pat, padds. Lat. pés, pedis. 

Eis, gen. ἑνός. In Cret. ἕνς and neut. ἕν the m of *sem- has become v 
according to phonetic law (see Brugmann Gr.‘ 86, 88). The weak 
form of the stem appears (*sm-) in ἅμ-α (simul), ἅ-παξ (Skt. sa-kft) and 
(*sm-) in p-ia and μ-ῶνυξ. 

φρήν, ppev-ds. Derivation uncertain. For conjectures see Boisacq. 

Zeus (see above, p. 142). 

Χιών originally an m-stem (Brugmann Gr. 88), cf. Lat. hiem-s. 

Bots, Skt. gaus. The weakest grade of the stem *g¥u- is seen in 
ἑκατόμ-βη (cf. Skt. sata-gu-=having 100 kine) and possibly in βόσ-πορος. 

(b) To the second class belong : 

Ὗς (Lat. sus), ἰχθύς, ὀφρύς, ὀσφύς (for accentuation see above, p. 141), 
ναῦς (from *vaus, Skt. ndd-3. See above, p. 142), χείρ (p. 138), ἅλς 
(p. 132) (Lat. sal), θρίξ (p. 130). 

(For full treatment see Brugmann Grd.? τι. i. 130-146.) 


FORMATIVE SUFFIXES. 


§ 133. Whether the suffixes which play so 
important a part in the stem formation of Greek 
nouns ever had a separate existence im the prehistoric stage of 
the original Indo-European language we cannot say. In O.E., 
dém meaning judgment was a separate noun, though it also 
appeared as a suffix in such words as cynedém (kingdom), freo- 
ἄδην (freedom). In the same way Ὁ. Εἰ. lic, which meant body 
or corpse (as preserved in the words lych-gate and lyke-wake) 
became the very common suffix -ly, so that manly or man-like 
means “having the body or form of a man.”1 Analogy is a 
tempting form of argument, but in the absence of convincing 
evidence, we must leave the question of the sometime separate 
existence of the Greek suffixes unanswered, and confine our 
attention to their function in the historical period of the lan- 
guage. The Indian grammarians distinguished between prim- 
ary and secondary suffixes used in noun formation, primary being 
those attached to a root or to a verbal stem, secondary those 
added to a noun stem. Thus -top- in δώ-τορ-ες was primary, 


1 See Giles”, 246 ff. 


Suffixes. 


334 A GRAMMAR OF NEW TESTAMENT GREEK. [§ 133-134 


τίο- in πάτρ-ιο-ς secondary. Modern philologists (e.g. Giles?, 
245) apply these terms differently, using primary of a suffix 
which is added directly to a root, and secondary of a suffix 
which is added to an existing stem which already contains a 
suffix. Thus -vo- in éapi-vd-s is secondary because it follows 
the locative case ending of the noun. For our treatment of the 
subject, however, the distinction may be ignored for reasons 
given by Brugmann (Grd.? τι. i. 124 f.). 


A. Vowel Suffixes.! 
(1) -o- and -a-. 


1. Nouns τη -os. 


§ 134. This -o- was originally the second syllable in 
dissyllabic light bases, and in the primitive I.E. period 
served a noun function, e.g. AvKo-s (Skt. vfka-s)<I.E. *ulq¥6-8; ζυγό-ν 
(Skt. yugd-m); also adjectivally, e.g. vé(F)os (Skt. nava-s). ‘The ablaut 
relation o/e appears in the vocative λύκε (Skt. vfka, Lat. lupe). The 
suffix -o- was then attached to bases which had the accent originally on 
the first syllable. In the same way the -a-, originally the second syllable 
of dissyllabic heavy bases, appears to have been extracted and given an 
extended use as a feminine suffix in the I.E. period. 

The most important use of the -o- suffix is in connexion with verbal 
roots. Two classes of nouns are thus formed, according to the accent ; 
the barytones are nomina actionis, and the oxytones nomina agentis. 


"ος 


(a) Nomina actionis are γόμος, δρόμος, πλό(ξ)ος, πόνος (πένομαι), 
σπόρος, τόκος, τρόμος, τρόπος, φόβος (φέβομαι), φόνος, φόρος : and with 
changed meaning λόγος, νόμος, ὄγκος (ἐνεγκ-εῖν), τάφος (θάπτω), τοῖχος. 

(b) Nomina agentis are τροφός, τροχός. 

Both classes became important in providing the latter member for 
compound nouns and adjectives. The NT provides examples in χιλί-αρχος, 
ἀρχ-ηγός, πρό-δρομος, οἰκο-δόμος, δεξιο-λάβος (δεξιο-βόλος, Lin Ac 2338), 
ἔμ-πορος, ὁδοι-πόρος ( > ὁδοιπορέω), ἱερό-συλος. 

For the accentuation of such compounds see below, p. 392 N.2. 

Compounds, not only from verbal stems, but from nouns with various 
stem-endings, often took this suffix. See below, Adjectives in -os. 


2. Nouns 1n -a, -ἡ. 
Verbal abstract nouns in -ἄ (-7), formed directly from the 


1 An obelus (1) before a word in the chapter on Suffixes marks it as 
not appearing before Aristotle. 


§ 134) SUFFIXES. 300 


root, were inherited at an early stage, and were still in active 
formation during the historical period : 

ἃ {τὸ E.g. ἀρχή, Body (βαλ-εῖν), δίκη (cf. δείκ-νυ-μι, Skt. 

dig-, dik, Lat. dico and judex (*tous-dic-s)), δοχή 

(δέχ-ομαι), κλοπή (ἔκλεπ-.ω), νομή (νέμ-ω), πνοή (πνέω), ῥιπή (ῥίπτω « 

ἘΡρῖπ-ιω), ῥοπή (pera; for ῥιπή, 1 Co 15°* DG), σπορά (σπερ- or σπαρ-): 

σπουδή (σπεύδ-ω), στεγή (στέγω, cf. Skt. sthagati, Lat. tego), στολή (στέλλω), 

σφαγή (σφάζω « ἔσφαγ-,ω), ταφή (θάπτω, ταφ-εῖν), τροπή (τρέπω), τροφή 
(τρέφω), φθορά (φθείρω «-φθερ-(ω), φυγή (φυγ-εῖν), χαρά (χαρ-ῆναι). 

Norr.—The transition from the abstract to the concrete is specially 

noticeable in ἀκοή, which is often used for the organ of hearing as well 


as for the object heard, and γραφή (see Vocab. s.v. for its common use in 
papyri as “ document of contract,” and its use in Aristeas and NT for 
“ Scripture ”’). 


From the classical age all kinds of verbs gave rise to such 
formations, especially those ending in mutes, and more particu- 
larly those whose stems ended in mediae or aspirates, e.g. 
ἁρπαγή. διδαχή, καταλλαγή, παραλλαγή, ταραχή. 

A few back-formations in -ἢ are of special interest. 

Τ᾿ Αγάπη, a shortened form of ἀγάπησις, just as συναντή (3 K 1818, 
4 K 215 58) and ἀπαντή (3 K 2018) are back-formations from συνάντησις 
and ἀπάντησις. (See Vocab. s.v. and supplementary note in Hap T 
xxvi. 139.) 

Τοἰκοδομή, shortened in the same way from οἰκοδόμησις, appears first 
in Aristotle, and is frequently found in the Κοινή, both literary (see 


Lobeck Phryn. 487 ff.) and vernacular (see Vocab. s.v.), for οἰκοδόμημα ; 
-a- is also used to represent the feminine equivalents to masculines in 


τος : 6.6. θεός : θεά, ἀδελφός : ἀδελφή. In this the adjectival type -os, 
-a, -ov was followed. 


3. Adjectives in -os. 

(a) Simplicia: ἀγαθός, βάρβαρος, ἐνεός (““ étym. inconnue,” Boisacq), 
ἔρημος, κωφός, λεῖος, λοιπός, μάρμαρος, μόνος, μωρός, νόθος, ὀλίγος, 
πειθός, σοφός, στενός, φίλος, χαλεπός, χωλός. 

(6) Compounds: ἄναλος, πρόγονος, ἀργός (see pp. 158, 287), δύσκολος, 
ἔγκυος, ὅμιλος (see below, n. 1), ἄπειρος, ταλαίπωρος, Τἄραφος (for ἄραπτος, 
see p. 371), ἄσπονδος, Ττρίστεγος, δίστομος, ἄστοργος, ἱερόσυλος, 
αὐτόφωρος. 

ΝΟΤΕ.---]. ὅμιλος (which owes its place in NT text (Rev 1817) to 
cursive 1, the sole MS used by Erasmus for the Apoc.—a rdg. also found 
in Hipp.) is given a note here because it has not been included in the 
ὁμο- epds. in ὃ 107 (p. 284). Siitterlin 61 suspects that itis acpd. Boisacq 


336 A GRAMMAR OF NEW TESTAMENT GREEK. [§ 134-135 


(p. 700) derives from ἔόμο-μῖλ- by syllabic haplology and eft. Skt. mildti 
to come together, join, méla-h meeting, Lat. miles -itis (mil-it-és ““ *ceux qui 
marchent en trowpe’’). 

2. Adjectives formed from composita generally took the suffix most 
appropriate to the second member of the epd. But when the first 
member was a particle or a preposition an -a- stem gave way to the suffix 
-o-. This usage spread to most cpds. ending in -a- stems, and consonantal 
stems also showed a strong preference for the convenient suffix -os. See 
Debrunner Wortb. 71 ff., and p. 289 above. 


(2) -co- and -ιᾶ-. 


1. From adverbs and locatives in -t- we have— 


ἐπεὶ § 135. ἼἌρτιος (ἄρτι), πρώιος (πρωΐ and 80 by 
analogy ὄψιος (ὀψέ), ἐνάλιος (ἐν ἁλί) and παράλιος, 
δεξιός (cf. δεξι-τερός). So ἀΐδιος (for the locative ἀεί see Vocab. 8.0υ.). 
Under this head should also come μέσος (cf. Lat. medius, Skt. madhyas. 
Brugm. Grd.? τι. i. 164 relates this to *me-dhi, με-τά and Skt. d-dhi) and 
ἄλλος (*alio-s, Lat. alius, from *ali, cf. Lat. ali-ter). 


2. Verbal adjectives. 


Eg. ἅγιος (ἄζομαι, Skt. yajya-s), ἄξιος (*ay-ri-os, Boisacq, s.v.), σφάγιος 
(:- σφάγιον, σφάζω -- σφαγιω). In this way comes μανία (μαίνομαι --1.Ἐ]. 
*mni-6-). 

In I.E. there were a number of adjectives in (ι) so with verbal force (see 
Brugm. Grd.? 11. i. 183 ff.), and this tendency has not been without effect in 
Greek, though the suffix did not prove a fruitful source for adjectives. 


3. Denominative adjectives—This very numerous class con- 
sists of adjectives formed from the stems of nouns by means of 
the suffix -ἰος with the meaning “ of, or belonging to.” 


Thus fSoktpros (-“δοκιμή «--δόκιμος «- δέκομαι), τίμιος (-««τιμή), κόσμιος 
(<koopos). 

The suffix is thinly disguised by contraction in θεῖος : θεός, meLos 
(πεδ-ιος, cf. Skt. padyds) : πούς. 

This came to be the appropriate adjectival suffix in epds., e.g. 
καταχθόνιος, μακροχρόνιος. 

From such regular formations as σωτήριος the new 
suffix -τηριο- was extracted, which appears in 
ἱκετήριος, and in the neuter nouns in -τήριον given in ὃ 137 below. Apart 
from a rather numerous list of derived substantives (represented in NT by 
teuxtypia and σωτηρία), this suffix soon worked itself out, and was 
displaced by -τικός. Thus we have κριτήριον, but κριτικός (see p. 379). 


-τήριος 


§ 135] SUFFIXES. 337 


The addition of -cos to various -t- stems produced 
-σιος (K. BI. ii. 292), ἑκούσιος : ἑκοντ-. In this way 
-σιος came to be the regular adjectival formation corresponding to the 
verbal adjectives in -ros, 6.9. θαυμάσιος : θαυμαστός, and to the denomin- 
ative nouns and deverbativa in -rns, e.g. δημόσιος : δημότης (-: δῆμος), 
γνήσιος : γνησιότης (-«γίγνομαι). 

ΝΟΤΕ.---Ἰ. Debrunner Worth. 143) observes that this -σιος has 
become specially attached to compounds, since the verbal adjectives in 
-ros favoured the compounds, and the nomina agentis in -rns were origin- 
ally also peculiar to the compounds ; in particular the fem. noun in -cia 
thus came into sharp contrast with -o.s in the simplex verb. Thus 
γνῶσις : ἀγνωσία, δόσις : Γμισθαποδοσία, κρίσις : Τδικαιοκρισία, στάσις : 
ἡ ἀποστασία. 

2. Φιλιππήσιος (Ph 415) does not, of course, belong to this group, but 
has the Gr. -cos added to the -e(n)s- stem of the Lat. -ensis. For the loss 
of -n-, see above, p. 106. Ramsay (J7'S i. 116) says this suffix is only used 
in Gr. to reproduce a Lat. name, as Μουτουνήσιος for Mutinensis. The 
ordinary Gr. was Φιλιππεύς or -nvds. Possibly the remembrance of 
Homeric Ἰθακήσιος (Bl°-D 6 n. 6) may be a factor. 


=CLOS 


With a- stems -cos combines to form -atos (rarely 
-aos as δίκη : δίκαιος) ; thus we have ἀγοραῖος, 
ἀναγκαῖος, ἀρχαῖος, Τἀκρογωνιαῖος (see ὃ 106 (a) and Vocab.), ἑδραῖος, 
κρυφαῖος, σπουδαῖος, ὡραῖος. 

Brugmann (Gird.? 1. i. 194) finds the explanation in the contraction 
of *-acw- to -ao- in the three words κνεφαῖος, γεραιός, γηραιός (the 
accent in the last two following that in παλαιός). The rest would then 
be formed by false analogy. Brugmann’s earlier explanation (Gr.* 181, 
retained by Thumb in Gir.* 212) rests on the loc. sing. -av-+-so-. At 
any rate the locative meaning so evident in ἀγοραῖος, θυραῖος and 
πυλαῖος gave the meaning “* belonging to a place” to -atos, and it was 
added in this sense to o- stems. Thus in class. Gr. we find νησαῖος and 
xepoaios. This predominant meaning may account for its wide use in 
gentilic names. In addition to those given above on p. 150, the NT 
supplies ᾿Αθηναῖος, Βεροιαῖος, AepBatos, Κυρηναῖος. For δευτεραῖος 
and τεταρταῖος see p. 176 above (also W. Bauer HNT ap. Jn 1139). 


-atos 


In παλαιός and Kpatatds the suffix -os is added 


=A.L0 ; 
: to the adverb πάλαι and to κραται (cf. κραταί- πους). 


Like δίκαιος (δίκη) the following are proparoxy- 
tone : βίαιος ( : Bia), μάταιος (< μάτην < parn), 
βέβαιος (<Baive, οἵ. βέβα-μεν), ἀκέραιος (accented as almost all com- 
pounds with d-). For the noun προσκεφάλαιον see above, ὃ 126 (0). 


παιος 


Ταλαντιαῖος represents a class of adjective in 
which the suffix -ἰαῖος stands for measure, weight or 
value. Kiihner-Bl. ii. 292 f. suggests an origin in -ἰ- stems, σταδιαῖος, 
ὀργυιαῖος, though these first appear in Hellenistic. Μηνιαῖος is early ; 


“ταῖος 


338 A GRAMMAR OF NEW TESTAMENT GREEK. [8 135-136 


for its frequent use in papyri of monthly accounts see Mayser Gr. i. 448, 
and add Wilcken Archiv ii. 126 and P Oxy xii. p. 48. 
The suffix -evos represents the convergence of two 


=€LOS Ξ 
formations. 


(a) -ecos<*-eovos, attached to s- stems. 

Eg. τέλειος : redos (τελεσ-ιος), ἐπιτήδειος (<ady. ἐπιτηδές, see above, 
ὃ 120), "Apetos : Ἄρης. 

(Ὁ) -nios (Hom.) <*nF-.os attached to stems in ev-: v. 

In this way arose βασίλειος ( : βασιλεύς), ἀστεῖος ( : ἄστυ). 

Oiketos ( : οἰκεύς, Hom.) would be popularly attached to οἶκος after 
oikevs had become obsolete. With the help of this analogy the rapid 
extension of the suffix to other stems can be explained. 

᾿Ανθρώπειος has its natural counterpart in γυναικεῖος. So also we find 
αἴγειος ( : ai€), Τἐπάρχειος (ἔπαρχος), ἐπίγειος (see above, ὃ 120), weyadetos, 
ἀχρεῖος (for deriv. see above, ὃ 107, also Boisacq 1070). 

This suffix, according to Brugmann (Gr.‘ 212), 
originates in loc. sing. in -o.-+yo and is found in 
ποῖος, τοῖος, ἀλλοῖος etc. with παντοῖος conforming to type. 

It is represented in NT by ὁποῖος and ὅμοιος. 

The locative derivation is disputed by Hirt Hdb. 255. Debrunner 
Wortb. 144 derives the suffix from *o-o.Fos, and regards it as originally a 
Kompositionshinterglied with the meaning “‘ kind,” “ sort.” 

Two examples in NT. 
Natpgos (in Hom., Hes. and Herod. always 
πατρώιος ; see LS. From stem πατρωξ-. Brugmann Gird.? τι. i. 206). 

Ὑπερῷος (an adjectival ending added to ὑπέρ, following πατρῷος, 

untp@os), represented by the neut. noun ὕπερῷον. See p. 326. 


ποῖος 


πῷος 


(3) Nouns in -ia. 


§ 136. Several important groups call for con- 
sideration. 

1. Names of Countries—Many of these are back-formations from 
national names. Συρία < Σύριος « Σύρος; ᾿Αχαΐα « ᾿Αχαιός; Φρυγία « 
Φρύγιος << Φρύξ will serve as examples of many more in NT. 

2. Abstract Nouns.—The normal type is found in ἐλευθερία 
(«ἐλευθέριος <eevbepos), ξενία (-- ξένιος -- ξένος), σωτηρία (<carnpios< 
σωτήρ). So with ἡσυχία, παρθενία, σκοτία, φιλία. 

But a far larger number are formed straight from nouns, or from 
adjectives in -os without the intervention of an adj. in -vos, e.g. : 

from adjectives in -os, δειλία, ἐλαφρία, κοινωνία, μαλακία, μωρία, 

πικρία, πονηρία, ῥυπαρία, σοφία ; 

from nouns, ἀγγελία (ἄγγελος), ἡγεμονία (ἡγεμών), ἡλικία (FAL), 

κυβία (κύβος), μαγία (udyos) ; 

from verbs, ἁμαρτία (“άνω, ἁμαρτ-εῖν), μαρτυρία (“έω) ; 


, 
“τα 


§ 136) SUFFIXES. 339 


from compounds, e.g. [ἐπιχορηγία (“yéw), {κενοδοξία ({κενόδοξος). 
So ἀγρυπνία, Τἀντιμισθία, ἐπαρχ(ε)ία, Τέἐφημερία, φιλαδελφία, 
(§ 108), φιλανθρωπία, φιλαργυρία, φιλονεικία. 
Compounds in -ἰὰ are formed direct, 6.5. ἀναλογία (from prepositional 
phrase, see ὃ 113), Τἐθελοθρησκία (ὃ 108). 


(4) Nouns in -ela. 


Abstracts in -e¢a are formed either from 
nouns in -evs (originally through the mediation 
of an adjective in -evos, for -ἤϊος), or from verbs in -εύω. 


, 
=ELa 


Thus βασιλεία « βασίλειος « βασιλεύς. 

All the other abstracts in -εία found in the NT are derived from 
verbs in -evw (-evopar): viz. ἀλαζονεία, dpeoxela*, δουλεία, epibeia*, 
Epnveia*, θεραπεία, TOpyoxela, tepateta*, κολακεία ἢ, λατρεία, Τλογεία, 
tpeOodeia*, μοιχεία, νηστεία, Τοἰκετεία, παιδεία, ἱπερισσεία, πολιτεία, 
πορεία, πορνεία, πραγματεία, πρεσβεία, Ἱπροφητεία, στρατεία. φαρ- 
μακεία, φυτεία. ‘ 

Norr.—1. Those nouns marked above with an asterisk (*) are spelt 
by WH in the shortened form -ia (App.? 161). This is due to the itacistic 
tendencies of the scribes in the age of the great uncials. See above, 
pp. 57, 76 f. and Proleg. 47. 

2. The transition from abstract to result is seen in some words, e.g. 
φυτεία planting in Xen., LXX, but plant in Mt 15% (as in inserr.) ; to 
collective concrete in θεραπεία (Lk 1242), οἰκετεία (Mt 2445), πρεσβεία 
(Lk 1432 1914), Θρησκεία hovers between worship and the ritual of 
worship (see Vocab.). Μεθοδεία used in malam partem in Eph 414 64, 
as in Polyb. and LXX. See J. A. Robinson in loc. Vocab. s.v. shows 
that in papp. the word reverted in meaning to method. 

3. For ἀρεσκεία and ἐριθεία see p. 57. That the former is derived 
from ἀρεσκεύω is clear from the bad sense in which it is used by class. 
writers and in literary Hellenistic. For vernacular support of Philo’s 
use in good sense, as in NT, see Vocab, ᾿Ἐριθεία «ἐριθεύομαι to work for 
hire, <épios day labourer, worker for hire. Selfishness rather than 
factiousness was the original meaning, labouring for one’s own interests 
rather than devotion to public service. So in the 3 NT passages. See 
Kennedy EGT ap. Phil 115, and Vocab. 

4. The word \oyela was rescued by Deissmann (BS 142 ff., 219f., 
LAE 103 ff., 2104) from Grimm’s class of “‘ biblical words,” and the dis- 
covery of its verb λογεύω to collect (see Wilcken Ostr. i. 255 n.1, 493 f.) 
removes its derivation from doubt. For further instances of this word, so 
common in the papp. and yet entirely absent from literary sources, see 
Vocab. s.v. Προφητεία is Hellenistic, with very slight support until we 
come to papp. and inscrr., for which see Vocab. s.v. 


340 A GRAMMAR OF NEW TESTAMENT GREEK. [§ 136-137 


Masculine nouns in -ίας originate from 


, 
“Las . = 
abstracts in -a. 


Thus veavias, the only NT member of this class, is derived, acc. to 
Debrunner (Worth. 145), from *veavia youth (abstract), which then gave 
youth (concrete collective) from which the individualised masc. form in 
-las came. 


(5) Nouns in -o-ia. 


; Abstracts in -ova are derived in the same 
“ola 5 . τος ἐς * : 
way as the adjectives in -ovos (Ἐ-τιος) ; see 


above, ὃ 135. 


(a) From composita came ἀγνωσία, ἀθανασία, ἀκαθαρσία, ἀκρασία, 
amotia,! Τἀκαταστασία, dowrtla,! Τἀφθαρσία, Ταἰχμαλωσία, ἵμισθαπο- 
δοσία, ἱπροσωπολημψία, Ταἱματεκχυσία, ὁρκωμοσία, νομο-, νου-, Tépo-, 
Τυΐϊοτθεσία, ἱκενοδοξία, Τδικαιοκρισία, Τόὀλιγοπιστία,, εὐεργεσία, 
εὐχαριστία, ᾿παλινγενεσία, πλεονεξία, ἵἹπρωτοκλισία, παρρησία, διχο- 
στασία, ἐκκλησία, μετοικεσία, Τ ἀποστασία, συνωμοσία. 

(b) The only simplicia to come under this heading are θυσία and 
κλισία. 

(c) Seven abstracts from verbs in -ἀζω (-ἀζομαι) complete the list: 
γυμνασία, δοκιμασία, ἐργασία, ἑτοιμασία, Τόπτασία, παραχειμασία, 
φαντασία. 

Norr.—1. Παραχειμασία, from “αἴω (see ὃ 123), a verb used by 
Polyb. ii. 54. 5. The simplex χειμασία is also used in the same context 
of winter quarters, upon which Capes remarks: ‘‘ The word is used by 
Herod. of the haunts of birds (ii. 22. 6), but only reappears in the later 
language.” 

2. For fuller treatment see Fraenkel ZVS xlv. 160 ff. 


(6) Neuter Nouns in -ἰον. 


§ 137. This extremely large class of nouns 
requires systematic treatment, both because 
of the common belief that its fundamental characteristic is the 
diminutive meaning, and on account of its prolific extension 
from the Hellenistic age onwards. It is necessary to separate 
nouns which do not originate with the diminutive connotation 
from those that do. 


=LOV 


1 For phonetic reasons the τ- remains unchanged in these words. 


§ 137] SUFFIXES. 341 


1. Nouns in -ἰον (other than diminutives). 


Here we may follow the same classification as in § 135 above 
when dealing with adjectives, for the substantivised neuter of 


such adjectives probably marks the starting-pomt of this 
development. 


1. Adverbs and locatives in -. do not seem to have contributed nouns 
to our NT vocabulary. 

2. Adjectives—(a) Verbal adjectives (as in I.E.) are responsible for 
a number of nomina actionis. In some of these the verbal force is strongly 
felt throughout the history of the word; in most the concrete meaning 
superseded the abstract at an early stage. Closely associated with these 
are many denominatives in -ἰο- themselves formed from nomina actionis 
and nomina agentis. Other nouns are formed directly from verbs without 
the intervention of any adj.: 6.9. ἵπροσφάγιον and Τσιτομέτριον. For 
extra-biblical use, see Vocab. s.vv. The noun may have come to represent 
the result of the action, as in λόγιον, εὐαγγέλιον, {θεμέλιον ; or its object, 
as σφάγιον ; or its instrument, as ἱμάτιον, Τὀψώνιον (see Vocab. s.v. for 
history and meaning), παραμύθιον. Sometimes the subjects of the 
action are expressed collectively, as in συμπόσιον, συνέδριον, Τσυμβούλιον 
(see Deissmann BS 238 and Vocab. s.v. for this late word), and by analogy, 
Τπρεσβυτέριον. 

This varying relation of the noun to the verbal action is sometimes 
reflected in the different uses of the same word. {γεώργιον in 1 Co 3° 
(a rare word, see J. Weiss in loc.) seems to =husbandry, tilth, as in LX X' ; 
whereas in Pr 245: 30. as in Strabo, it=field. So μαρτύριον. 

(b) Very near to these in meaning are adjectival abstracts in -tov 
expressing a state or attribute. 

The change in relation to verbal action may account partly for the 
two meanings of αἴτιον in Ac 1930 (cause) and in Lk 234. 14. 22 (=crime). 
So Petersen Gr. Dim. 27, “‘ τὸ αἴτιον ᾿ the cause,’ with the distinct idea 
of activity, must have been originally conceived as ἡ the blame-worthi- 
ness,’ for it comes from the adj. αἴτιος, ‘ blame-worthy.’ ” 

The transition from abstract to concrete is seen in δαιμόνιον, 
divinity (---δαιμόνιος, divine). From a vague meaning of “ divine power,” 
a personal denotation was evolved. The limitation in Attic (as in magic, 
see BS 281 and Vocab. s.v.) to inferior deities is probably due to influence 
of diminutives in -cov. See below, ὃ 138. 

(c) Compound adjectives. We may note specially two kinds: 

a. prepositional compounds, in some of which the adjective sur- 
vives, e.g. μεθόριον, ὑποζύγιον ; but not in all, eg. ἐνύπνιον 
(see above, ὃ 109), προαύλιον (§ 125 (c)). 

8. numerical compounds, and those signifying a part of the simplex. 
Cpds. of ἡμι- and ἀκρο- are common. In the NT we find 
Τἡμιώριον (Rev 8! ΝΡ 046), ἀκροθίνιον (§ 106), ὑμεσονύκτιον. 


342 A GRAMMAR OF NEW TESTAMENT GREEK. [ξ 137 


3. Denominatives. (As some of these are formed from nomina 
actionis they have been referred to under 2 (a) above.) 

The infinite variety of meanings connoted by this suffix forbids an 
exhaustive classification. The main groups come under the following 
headings : 

(1) Appurtenance. 

(a) Place. 

a. The primitive is nomen agentis. The derivative denotes the 
place connected with the person. ᾿ἘἘμπόριον (<éuropos), 
treNdviov (--τελώνης). (CE. δεκατώνης : δεκατώνιον iv/B.C.) 

A special group is formed from nom. agentis in -rnp, of the type 
δικαστήρ : δικαστήριον. The termination -τήριον was then detached and 
applied to nomina agentis in -rns. Thus Τἀκροατήριον, δεσμωτήριον, 
κριτήριον, οἰκητήριον (κατ-). 

8. The sanctuary of a god or hero. ᾿Απόλλων : ᾿Απολλώνιον, 
Νύμφη : Νυμφαῖον. So ᾿Ασταρτεῖον 1 Καὶ 3119.1 On this 
analogy Τ᾿ εἰδώλιον (an idol’s temple, 1 Co 819). 

(b) Instrument or Means. 

a. Sometimes the primitive is not extant, e.g. ἱμάτιον (<*iva 
<*Ficna< ,/*uis), ‘‘ that which is used for winding about 
or surrounding.” πηδάλιον (<anddv=blade of oar), 
Τγαζοφυλάκιον (see above, ὃ 105). 

8. Primitive is a nomen agentis in -rnp (-Tns). 


Αἰσθητήριον (see Vocab. s.v.), θυμιατήριον, θυσιαστήριον, Τἱλαστήριον 
(Deissmann BS 124 ff.), μυστήριον, σωτήριον, τεκμήριον (see Vocab. 
s.v.), φυλακτήριον (see below, ὃ 150), καυ(σ)τήριον (-«- Ἰκαυστηριάξω). 

(c) Vessels and utensils, household effects. 

These really form one congeneric group with words like ποτήριον, 
Τὑπολήνιον, ὑποπόδιον. The primitive of τρύβλιον is not extant. 

(4) Herd of domestic animals. E.g. ποίμνιον (=herd of sheep: ποιμήν). 
See below, p. 346. 

(6) Part of the whole designated by the primitive. E.g. κράνιον : κάρανον 
=that which belongs to the head, the skull. +Batov (palm branch : Bais).? 

(f) Indefinite plurals in -ta. “Opta=what belongs to the boundary, 
limits, frontier. opria=the things belonging to the load (-- φόρτος -Ξ 
load), then φορτίον --ἰοαά. Xwpia=what belongs to a particular country 
(χώρα), regions. Then χωρίον, a particular place, or, as in Mk 14%, an 
enclosed piece of ground (RVmg). 

(g) A word that comes under none of these headings is Τδυσεντέριον 
(Ac 28° NAB, see p. 125). Can this late form of the word be influenced 
by μεσεντέριον, one of “a large number of words beginning with μεσο- 
which are of a heterogeneous semantic character ’’ (Petersen, op. cit. 37) 1 
For its derivation see p. 287 above. 

1Cf. P Gurob 221° Μιθραίου, 22 ᾿Αφροσιδίου, 37 ἝἙρμαίου, 43 Σαχμιείου, 
49 Νεφθιμιείου (iii/B.C.). * The new LS accents βάϊον (Bdis). 


§ 137] SUFFIXES. 343 


; ΝΟΤΕ.---Ἰ. Under (a), (6) and (c) we see examples 

Gad a of -τήριον as a suffix. Ὁρμητήριον (found in Xen., 
Isocr. and Dem. and more freely in Hellenistic authors), πολεμητήριον 
(Polyb.), show that in Attic and literary Hellenistic the suffix was already 
productive in the formation of new words. For new words appearing in 
Ptolemaic papyrisee Mayser Gr. 1. 439 ἢ. For later words, cf. ἀπαντητήριον 
(#nn) in PST iii. 175° (a.D. 462). ᾿Αγνευτήριον occurs in the Gospel frag- 
ment P Oxy v. 840° 13, 

2. Some nouns in -τήριον pass from the local meaning in the later 
language, e.g. βουλευτήριον, council chamber in class. Gr., becomes senate 
in Polyb. ii. 50. 10; ef. Dion. H. 2. 12. 

Κριτήριον, which is used with both the instrumental and the local 
meanings in Plato (so papp. and inscrr.)=tribunal Jas 2°, but law-suit, 
cause 1 Co 6* 4 (see J. Weiss tn loc.). 

(2) Material, substance. 

This can be illustrated from the NT by ἀργύριον (ἄργυρος) silver, 
silver coin, money; χρυσίον (χρυσός) gold, piece of gold, money, golden 
ornament ; χαλκίον (χαλκός) brazen vessel (Mk 74); κεράμιον (κέραμος) 
earthenware vessel; βιβλίον (βίβλος, βύβλος) that which is made of 
papyrus pith, a papyrus roll, book ; σχοινίον rope, that which is made of 
reeds (σχοῖνοι) ; ὀθόνιον (ὀθόνη) that which is made of linen, linen cloth, 
bandage, swathing band; σιτίον (σῖτος wheat) grain (generally used in 
plur.=food, provisions); κηρίον (xnpds=bees-wax) honeycomb (Lk 2442, 
ΕΚΜΘ.. 

(3) Category. 

This class of words with the connotation ‘‘ belonging to the category 
of,” “‘ having the nature of,’ Petersen divides into those in which -.oy 
has a ‘‘ generalising’? nature, under which θηρίον would come, and 
those in which the suffix has a “ specialising”’ force. Under the latter 
heading must come σανδάλιον (from a root of Semitic origin) a shoe of the 
sandal kind; fretpd8.ov (see above, p. 176); ζῷον (which is adjectival 
in origin and has no noun primitive). 

(4) Similarity. 

NT examples are {kepdtvov (κέρας horn) carob-pod (Lk 1615), κολλούριον 
(p. 78) (κολλύραε κόλλιξ, a bread-roll) eye-salve, because put up in small 
cakes (see Vocab. s.v.), Ἱπτερύγιον (πτέρυξ) anything like a wing—in 
architecture, a turret, a pinnacle, battlement or pointed roof. 

+Tomdfiov, a word of uncertain derivation, in which the root is almost 
certainly Semitic (see Cheyne in HBi. s.v.), possibly owes its suffix to the 
influence of ὀνύχιον (ὄνυξ) “᾿ that which is like a nail,” “‘ a gem streaked 
with veins,” an onyx, ἀνθράκιον (ἄνθραξ) a kind of jewel that is like 
charcoal. Under this heading of Similarity come the names of quite 
a number of articles of dress and ornament. 

Under the influence of such words as δελφίνιον ( : δελφίς) the -ἰον 
suffix came to be used as the normal ending for plant names. In this 
way probably we can account for the two forms Τἀψίνθιον (Rev 8! Ν᾿ 


344 A GRAMMAR OF NEW TESTAMENT GREEK. [§ 137-138 


and ἄἅψινθος (Rev 81! A et al.); see above, p. 123. Here also should come 
in {ζιζάνιον (for its Semitic root see Lewy, Fremdworter, 52). 


As with adjectives in -vos, the nouns in -vov 
provide a subdivision in -eov. Here, again, 
we notice the convergence of two formations, those deriving 
from s- stems and those from nouns in -evs, or verbs in -evw 
(see above, p. 339). 


«εἴον 


(a) ᾿Αγγεῖον ( : ἄγγος). See Vocab. for variety of meaning in papyri. 
Δανεῖον ( : ddvos <*davds=Skt. dind-s. See Brgm. Gdr.? τι. i. 256, 526). 

(Ὁ) BpaBetov ( : βραβεύς), πανδοχεῖον (: πανδοχεύς), ταμεῖον 
(: ταμεύω) ; for spelling see above, p. 89. 

This suffix before long was appended to other stems, as we are re- 
minded by σημεῖον ( : σῆμα) and μνημεῖον ( : μνῆμα). 

Θεῖον (Hom. θέειον sulphur fumes) brimstone (<*6Feo-evov : <L.E. 
*dhuos breath); an adj. “‘ emitting vapours, fumes,” from a noun *6eds 
or Ἐθεόν (OFecds, -dv) ““ breath, fume.” So Boisacq 337; q.v. for con- 
nexion with θεός and θύω. 

Norr.—This termination was added to various stems in the classical 
age to betoken a place (especially for the carrying on of business), e.g. 
χαλκεύς : χαλκεῖον, κάπηλος : καπηλεῖον, Μοῦσα : Μουσεῖον. In Hellen- 
istic it proves a fertile source for such nouns as the papyri show, e.g. 
Τγραφεῖον (=“‘ writings” in Clem. ad Cor. 28°), Τἀγορονομεῖον and μνημονεῖον 
different names of the same notarial office (see Archiv vi. 104), ἀρχεῖον 
Τκαταλογεῖον. For all these terms see Mitteis, Papyruskunde, τι. i. 59 ff. 

A notable instance comes in a schedule of water rates, P Lond 1177 
(A.D. 113) (=iii. p. 183), where a Jewish place of prayer Τεὐχεῖον is 
mentioned in the same connexion as a mpocevyy or synagogue. 


2. Dimanutives in -ἰον. 

§ 138. This term is here used to include 
words with a “deteriorative”’ and “ hypo- 
coristic ᾿ 1 significance, as well as those which 
connote relative smallness of size. The entire class has probably 
arisen from the use of the suffix (see above, p. 343) to connote 
similarity to the type represented by the primitive noun. The 
original meaning is then “a kind of,” “‘ something like.” 


Diminutives 
in -.ov. 


Notre.—1. The question whether the diminutive suffix is inherited 
from I.E. or was developed in Greek within the class. period is still open. 


1The ancient Greek grammarians coined the word ὑποκοριστικά (from 
ὑποκορίζεσθαι --:““ to speak as a child (xépy)”’ ) for words formed to express tender- 
ness, and so smallness and even disparagement. It seems better to restrict 
its use to the original meaning. 


§ 138) SUFFLXES. 345 


Brugmann (Grd.? τι. i. 676) states that -(ἐ)ο- was a diminutive suffix 
in I.E. times. But his earlier view (Gr.* 180, retained by Thumb in 
ed.* 212) was that this development in the use of the suffix was post- 
Homeric. The absence of diminutives proper in Homer might of 
course be deliberate, but the total absence of “‘ faded diminutives”’ is 
significant. Two strong arguments in favour of the post-Homeric 
origin of diminutives are: (a) The dim. meaning is only found in the 
historic Gk. -vov, never in words such as πέζον (<zed-yov) or those in 
which the dissyllabic -ἰον is merged in a diphthong, as -aov, -evov, -ovov ; 
thus this function may be presumed to have developed after -dy-, -ασι-, 
-εσι-, τοσι-, had become -¢-, -a-, -εἰ-, -o-. (b) The accentuation of 
παιδίον, which was the principal pattern of diminutives and one of the 
oldest. Παῖς «πάξις was still often dissyl. in Homer (πάϊς), and the 
diminutive, if formed in early Homeric times, would be accented on the 
antepenult (παΐδιον >aidior). This suggests that παιδίον and all words 
modelled thereupon must be later at least than the earlier parts of 
Homer. See Petersen Greek Diminutives in -ἰον, from which this account 
is abridged. 

2. The accentuation of “ diminutives”’ seems to defy precise de- 
finition. Petersen (ib. 12 ff.) suggests “* trisyllabic substantives in -vov, 
if all connexion with the adjectival types from which they are derived 
has faded from the mind, have a tendency to accent the penult if they 
are dactylic, but the antepenult if they are tribrachs.’ But in view of 
the conflicting analogical influences at work both in early and later times 
he cannot make use of accent in his treatment of the semantic develop- 
ment of the -ἰον diminutives. 

Diminutives, which are found sparsely in tragedy, abound, as might 
be expected, in comedy. During the Hellenistic period the intimate 
language of the home and the speech of the lower social strata forced its 
way up into the vocabulary of literature. Words which originated in the 
nursery and the streets lost their hypocoristic meaning or became “ faded 
diminutives,’ in many cases superseding the primitive noun. (See 
p. 346, Note.) 


In the NT the following classes of diminutives in -vov are 
represented : 


Persons.—NauB8lov (παῖς), θυγάτριον ( : θυγάτηρ). tTekviov (1 Jn passim) 
( : τέκνον), κοράσιον ( : κόρη). In all of these the hypocoristic use is 
specially evident in the vocative. 

Animals.—Xtpov8iov (: στρουθός), προβάτιον ( : πρόβατον), ἀρνίον 
( : ἀρήν), νοσσίον ( : νοσσύ»). 

Parts of the δοάψ.---ἰ Ὡτίον ( : οὖς). 

Geographical terms.—{Nyotov ( : νῆσος). 

Food.—}Wryiov (=Wié) ἁ.λ. in Mk 738 (|| Mt 1527), ἡψωμίον Jn 1376. 
(elsewhere Marcus vii. 3, Diog. L. vi. 37 and papp.). By this time παιδίον 


346 A GRAMMAR OF NEW TESTAMENT GREEK. [ 138~139 


and προβάτιον are faded diminutives, for no consideration of size 
distinguishes them from their primitives. 

Ποίμνιον might seem to show a survival of hypocoristic use in Lk 12%, 
but the addition of τὸ μικρόν proves that there is no consciousness of 
diminutive force (see p. 342). 

*Qtiov is a faded hypocorism and has “‘ passed from the language of 
nurses and lovers into universal use. Both ὠτίον and ὠτάριον are never 
used when the ear is thought of as an organ of hearing, nor in a figurative 
sense, but only when the outer ear is distinctly in mind” (Petersen 7b. 
183). 

Norr.—Petersen cites Plut. Ages. 13 in proof that ἐπιστόλιον was 
a true diminutive. But its fairly common use in papyri without any 
such suggestion, e.g. P Fay 122° (c. 100 a.p.), P Lips 698 (118 a.D.), shows 
how easily the diminutive became a substitute for the primitive. Vogeser 
illustrates this tendency in the later vernacular and cites many passages 
where μικρόν is added (Gr. Heiligenleg. 42 f.). The full development of 
this tendency is seen in MGi. (See Thumb, Hellen. 178, 220, Handb. 62.) 


3. Conglutinates with -ἰον. 


§ 139. -ἰδιιον is a suffix which probably 
takes its rise from stems in -16-, e.g. ἀσπίδ-ιον 
(: ἀσπίς). The suffix was then applied to other stems, as 
in κλινίδιον ( : κλίνη). The uncertainty whether some words 
in -ίδιον had primitives in -0- or -ἰο- (e.g. In NT πινακίδιον 
«πινάκιον or «πίναξ may have given rise to the suffix -δίον 
which appears in ἰχθύδιον ( : ἐχθύς). 


«ἴδιον 


At the side of -ἴδιον there appears a suffix -ἴδιον, as in βιβλίδιον 
(-«----βιβλι-ίδιον ««βιβλίονῚ, οἰκίδιον (<*oixeidvov <oixia, but also assignable 
to οἶκος). In this way some have explained ἰχθύδιον : ἰχθύς. 

Further examples are ἀγρίδιον (Mart. Polyc. δ᾽), βιβλίδιον (Herm. 
Vis. iii. 14; Ignat. Eph. 201), ξιφίδιον (Mart. Polyc. 161), ἐπιστολίδιον 
(P Hamb 89 ; PST i. 9315). Thumb Dial. 376 quotes Pollux x. 116 to show 
that Aristoph. used Avyvidia for Avyva. Cf. P Rylii. 23971, ὀναρίδιον μικρόν. 
See also Mayser Gr. i. 428 for productivity of this suffix. 

Stems in -ap- provided a number of diminu- 
tives in -άριον which was then regarded as an 
independent suffix, and used to form a large number of diminutives, e.g. 
γυναικάριον ( : γυνή), κλινάριον ( : κλίνη), κυνάριον ( : κύων), ὀνάριον 
(: ὄνος), ὀψάριον (: ὄψον), παιδάριον (: παῖς), πλοιάριον (: πλοῖον), Τὠτάριον 
(+ ὠτίον). 

Notr.—1. The freedom with which this suffix was used in coining fresh 
words is seen from such a formation as κερβικάριον (Herm. Vis. iii. 1*) 
(<Lat. cervical <cervix), ξυλάριον (3 K 1713 and papp.; see Vocab. 434 f.), 
μοσχάριον (LXX and PST vi. 600*). 


-άριον 


§ 139-140] SUFFIXES. 347 


2. The decline and disappearance of the diminutive force in these -ἄριον 
formations becomes evident from their use in the papyri. Thus κοσμάριον 
(P Hamb 1043 χρυσᾶ ἐν κοσμαρίοις μναιαῖα), pepvapiov (BGU iv. 1102!9 where 
the amount of the dowry is quite substantial), ὠάρια (BGU iii. 781 v.*®). 
Οἰνάριον preserves a diminutive meaning in Epict. Hnch. 12 (ἐκχεῖται 
τὸ ἐλάδιον, κλέπτεται τὸ oivapiov), but the deteriorative force found in 
Demosth. ὁ. Lacr. 32 is retained in P Flor ii. 1603, according to Com- 
paretti’s note “a light wine, or of inferior quality,’ though there is 
nothing to indicate this in the context. See note in P Par p. 414 and 
additional ref. given in Witk.? p.27n. In MGr the process is complete, 
as λιοντάρι Shows (see Thumb Handb. 338). 


Not to be confused with these are the Latin loan words, in which 
assarius and denarius are given this common neuter ending, ἀσσάριον, 
δηνάριον, σουδάριον (sudarium, see BS 223), which, with σικάριος, is 
simply transliterated. For papp. see Vocab. s.v. For further creations 
under Latin influence see Vogeser, Gr. Heiligenleg. 41. 

These secondary suffixes can be yet further combined, as in Τβιβλαρ- 
ἴδιον (Rev 102: 88.), Τβιβλιδάριον (Rev 102 C, 108 x, 10° 046). Βιβλίδιον 
and βιβλάριον (Rev 10° A*) have also some MS attestation. In Herm. 
Vis. 11. 13, βιβλαρίδιον and βιβλίδιον are used together synonymously. 


(7) -εο-. 


The simple suffix -εος (<*-esos), which in Attic 
contracts to -οὖῦς (see above, pp. 120, 121, 156 for | 
irregularities in NT), forms a group of adjectives of material, represented 
by ἀργυροῦς, χρυσοῦς, χαλκοῦς, σιδηροῦς. The transition from material 
to colour, 6.9. πορφυροῦς, may have been helped by the secondary 
meaning of χρυσοῦς = golden. 

The old connexion between -ezo- and 7- stems, which is absent from these 
adjectives, appears in a few nouns which bear trace of adjectival origin, 
e.g. ὀστέον (cf. Skt. dsthi), and ὄρνεον (<dpys). In the same way the 
stem of στερεός appears in στέρεθος (Brugm. Grd.” τι. i. 199). 

Two masc. nouns in -eos seem to come in here. 

Oupeds in its Hom. meaning “ door-stone”’ shows its derivation from 
θύρα more clearly than in its Hellenistic meaning of “ shield’ (Polyb. 
Plut. Eph 615). 

Twheds (Mt 8° || Lk 958) occurs first in Arist. For derivation see 
Boisacq, s.v. 


-“εος 


(8) -ιἄ. 


8140. Widely productive at an early stage of the language 
for the formation of feminines to adjectives and nouns, this 
suffix now survives in longer suffixes in -a impure used for 

VOL. II. PART I1I.—24 


348 A GRAMMAR OF NEW TESTAMENT GREEK. [§ 140 


forming the feminine of third declension adjectives and parti- 
ciples with three terminations, e.g. -ovca (=-ovtia) etc. See 
§ 65 above. 


Φέροντ-, φέρουσα (-Ξεφερόντια), οὗ, Skt. bharant-, bharant-i. 
In γλῶσσα (=*yAw xia) the suffix is added directly to a root. 


1. Nouns in -evd. 


Feminine nouns in -ea are almost all 

regularly formed feminines of the eo-stem 

adjectives (<*-eo-1a). As a separate feminine was not used 
for these adjectives, the feminine was used substantivally. 


Ye 
“εια 


From adjectives in -ns we have ἀκρίβεια, ἀλήθεια, ἀναίδεια, 
αὐτάρκεια, συγγένεια, émetKera* (§ 120), εἰλικρίνειαῬ (ὃ 105), ἐνέργεια, 
κακοήθεια, συνήθεια, ἐγκράτεια, εὐλάβεια, d=, θεο-, εὐ-σέβεια, 
ἐπιμέλεια, κακοπάθεια, ξ΄ Ἱπραύπάθεια," ἀπείθεια, εὐπρέπεια, ἀσέλγεια 
(§ 107), ἀσθένεια, ἀσφάλεια (ὃ 107), συντέλεια (§ 127 (5)), Τἐκτένεια, 
ἐπιφάνεια, κατήφεια ; to these must be added ἀπώλεια, βοήθεια, 
ὠφέλεια 5, which are verbal derivatives. 

Three more which come from -es- stems have yielded to the influence 
of the -ia class, viz. ἀηδία (Lk 23! D, see Vocab. s.v.), ἀφειδία, εὐωδία. 
The same influence affected the orthography of other words. See p. 78 
above, also WH App.? 161, where the words marked * are spelt -ia, 
ἀκριβία, ἀπειθία (in Heb.) and éxtevia being treated as doubtful. 

Originally a similar suffix (<*-eF-,a or *-nF-,a) was used to form a few 
feminine titles to nouns in -εύς, whilst -eia (the fem. of adj. -εἰος) served 
for abstract nouns, e.g. βασιλεύς, king; βασίλεια, queen; βασιλεία, 
kingdom. But in the Tragedians -ἰς has already displaced -eva, and in 
Hellenistic the Macedonian suffix -ἰσσα established itself and spread 
widely from βασίλισσα to other feminine titles (see Mayser Gr. i. 255, 
and below, p. 349). 


2. Nouns in -αινα. 


This suffix, so largely used in Greek for 
feminines from -n- stems (*-n-1a@ >*-av-sa, 
>-awa), is only represented in NT by tydyypawa, the name 
Τρύφαινα, and the adj. μέλαινα. 


“αινα 


The origin of the suffix is seen in such a word as τέκτων : τέκταινα 
(=*rexrav-ra) (cf. Skt. taksan- : takSp-i). Then through similarity of 
nom. in -wv, this became fem. suffix for nouns with ovr- stem. The 
suffix was detached and appended even to -o stems, specially for names 
of persons and animals, e.g. λέων : λέαινα, λύκος : λύκαινα. 


§ 140] ~ SUFFIXES. 349 


Γάγγραινα may be derived from yoyypos (so Boisacq), or may come 
directly, with reduplication, from γραίνω = ypae, to gnaw. 


(9) -τρ-ια. 


TMaOyrpio is the only NT example of this 
form of the feminine of nomena agentis in -τῆς 
(<-tnp, see § 150). 


To Mayser’s instances (Gr. i. 444), βεβαιώτρια and προπωλήτρια, we 
may add ἀγοράστρια, P Thead 1" (a.D. 306), BGU iii. 9071" (ὁ. 185 A.D.). 


-τρια 


(10) -ισσα. 


This suffix (from ἔ-ἐκ-ια) arose with such 
examples as Κίλιξ : Κίλισσα, but was widely 
used in the Hellenistic period under the influence of the Mace- 
donian court. See (8) above. 


Ξισσα 


Βασίλισσα (see Vocab. s.v.), Φοίνισσα (Mk 725 Ὁ), Φοινίκισσα (ἰδ. 
Bet al.), Συροφοινίκισσα (ib. SAL e¢ al.). 

The papyri furnish further exx. of this fem. formation, e.g. πατρώνισσα 
IGSI, 1671, P Oxy iii. 47877 (a.D. 132), ἀρτοκόπισσα (new fem. of apro- 
κόπος) P Oxy viii. 11468-° (early iv/a.D.). Mayser (i. 255, 451) cites 
ἱέρισσα from numerous Ptolemaic papp., and the two adjectives μελανο- 
σπαλάκισσα (iron-grey) and ψακάδισσα (dappled) from a register of 
cavalry horses, P Petr ii. 35, col. 17: ὃ (ο. 240 B.c.). 


(11) -ἰσσος. 


The proper name Νάρκισσος preserves a plant 
name showing the same suffix as κυπάρισσος. See 
Vocab. s.v., and for the suffix -σσος, Kretschmer ἔτη. 405 ff. 


“LOGOS 


(12) -ev-. 


(a) From the earliest period this was freely 
used as a denominative suffix to indicate a 
person specially concerned with the thing denoted by the 
primitive. It thus came in time to mark a trade or profession. 


, 
=EUS 


Our examples are ἁλιεύς, Τβυρσεύς, Tyvades (see p. 108), γραμματεύς, 
ἱερεύς, ἱππεύς, κεραμεύς, φαρμακεύς (only in inferior MSS of Rev 218), 
χαλκεύς. 

(b) In a few words we have trace of an early formation from a verb, 
γονεύς (--γέν-εσθαι) and φονεύς (< J/oev. cf. Hom. ἔ-πε-φν-ε). 


350 A GRAMMAR OF NEW TESTAMENT GREEK. [§ 140-141 


(c) In compound formations this suffix is almost confined to pre- 
positional compounds (for reasons shown by Debrunner Wortb. 152 f.). 
ΤΚαταγγελεύς is a NT example. Πανδοχεύς ἰ5 an exception. Συγγενεῦσιν 
as read in some MSS in Mk 6*, Lk 2%, is really from the adjective cvyyevns— 
an instance of popular heteroclisis, see above, p. 138. 

(4) Ethnica were generally formed by substituting -evs for suffix in 
name of city or country, ¢.g. Θεσσαλονικ-εύς (-n), Κολόσσ-εύς (-αἱ), 
Λαοδικ-εύς (-ia), Ταρσ-εύς (-ds). 

From names in -ed, both -eevds and -evs (-ειεύς >-eevs >-evs) 
were formed in Attic, e.g. Δεκέλεια, Δεκελειεύς and Δεκελεύς ; in Hellen- 
istic almost always -evs, e.g. ᾿Αλεξανδρεύς, ᾿Αντιοχεύς. 


(6) Νηρεύς (Rom 16"), a personal name, possibly marking one of Nero’s 
freedmen ; see Rouffiac 91 for inscriptional evidence of widespread occur- 
rence in Galatia and Athens. The name, of course, is as old as Homer. 


(13) -ew- : -u-. 


(a) -ew is represented by the solitary noun πῆχυς 
(p. 141) and the adjectives on p. 160 above. 
(6) -u-. For nouns see p. 141. 


-υς 


B. Nasal Suffixes. 
(1) no:, “na 
1. Nouns in -μός, -θμός, -σμός. 


δ 141. This suffix originates in an LE. 
abstract suffix which appears in different 
stages of development in Gr. While it is used primarily to 
form verbal abstracts, that meaning is often lost in the concrete. 


-μός 


(a) In λιμός hunger, λοιμός plague, the underlying root has vanished, 
though both may perhaps be related to Lat. /étwm (=death ). 

(6) It is attached to primary verbal stems in a number of words, 
such as ἅρμός (probably from αἱ seen in ἀραρίσκω), βρυγμός (βρύχω), 
βωμός (<fa-, cf ἔ-βη-ν), ToToypappds (γράφω), διωγμός (διώκω), [ἐλεγμός 
(ἐλέγχω, also ἀπελεγμός fr. ἀπ΄), ὀδυρμός (ὀδύρομαι), οἰκτιρμός 
(οἰκτείρω), Τέἐμπαιγμός (ἐμπαίζω), σεισμός (σείω, cf. σέ-σεισ-μαι, ἐ-σείσ-θηνῚ, 
φραγμός (φράσσω), ψαλμός (ψάλλω). 

(c) Its more typical use is with denominative verbal stems, e.g. 
Ἰἁάρπαγμός, στεναγμός, στηριγμός. 

(4) Far more usual is the extended form of the 
suffix, -σμός, which is specially attached to the 
denominatives in -ἰζω, -a¢ and v¢w, with dental stems. 


-σμός 


δ 141] SUFFIXES. 351 


From -i¢@ come Τάγνισμός, -αρτισμός (Τάἀπ-, κατ’), Τβαπτισμός, 
βασανισμός (Alexis in Athen., 4 Macc), θερισμός, tipatiopds (see 
Vocab.), Τ᾽ Ιουδαϊσμός, Τκαθαρισμός, λογισμός, μακαρισμός, μερισμός 
(δια-), Τόνειδισμός, ἱπαροργισμός, ἵπορισμός, Τῥαντισμός, Τσαββατισμός 
(see Moffatt JCC, ap. Heb 49), ἐπισιτισμός (from Xen. down), ᾿ἀφανισμός, 
Τσωφρονισμός, Τφωτισμός, χρηματισμός, Ἱψιθυρισμός. 

From dé, ἁγιασμός, ἀσπασμός, Τἐνταφιασμός. πειρασμός, TTapa- 
πικρασμός. 

From -ύζω, Τγογγυσμός (see Rutherford NP 463), κατακλυσμός (see 
Vocab.). 

Then the suffix passed over to nasal and other verbs whose perf. pass. 
ended in -σμαι as with the - ἕω verbs, e.g. ἱλασμός (ἱλάσκω, cf. ἱλάσ-θητι), 
ἵμιασμός (μιαίν ω), ἡμολυσμός (μολύνω), παροξυσμός (παροξύνω), Τκυλισμός 
(2 Pet 2535 ΒΟ) (κυλίνδω, later form κυλίω, p. 246). 

Δεσμός exhibits this suffix as early as Homer (δέω, d¢-Se-par). 

Norr.—For papyrus nouns in -μός, see Mayser Gr. i. 435 ff., and for 
the very large class of new formations in -ἰσμός οἵ, δειγματισμός PSI 
iv. 358°: 21; ἐμπυρισμός PST iv. 3387. 15 (ν, Intr.), 3397 (all iii/B.c.). 

(e) In the termination -θμός we have the root 
determinative ! -dh- combined with the suffix -pos. 

᾿Αριθμός where the root is ἀρι-, as found in νήριτος (=numberless), 
TBabpds (see p- 112, Vocab. s.v.) from the root Ba- (seen in Baiva « ἔβαμιω 
<*q"m-Zo) By (ἔ-βη), κλαυθμός from κλαίω (fut. κλαύ-σ-ομαι). 


-θμός. 


(f) Several nouns in -yos go back to a very early 
stage in the language. 

“Avepos (I.E. andi, to breathe, cf. Skt. dni-ti). 

Δῆμος ( / da(t), cf. Skt. dati, to cut, divide) originally meant the portion 
of territory belonging to a community, then a canton, finally the people 
(see Boisacq 182). 

Κάλαμος (I.E. *k°l’mo-s, see Boisacg 397). 

Képapos (I.E. *ger?mo-s, cf. Lat. crémo from ν' *ger or *gar, cf. Lat. 
carbo (*car-dho), see Boisacg 436). 

Κόσμος (<*kovopo-s I.E. *kens-=to make authoritative announcement, 
cf. Lat. censeo=to judge ; see Boisacq 500). 

K@pos (opinion is divided between I.E. *q6(i)mo-s from a supposed 

J qoi=to meet, suggested as root of κώμη, and I.E. *kd[i]-mo-s, cf. Skt. 
gicati =to share with one, Boisacq 544). 
Μῶμος probably connected with ἀμύμων, for which Hirt suggests 1.10, 


πμος 


1 For root-determinatives see Brugmann KVG 296 f., Hirt Handb., 202 fi., 
Debrunner Wortb. 3f. The term is applied to a consonant coming between 
the root (or base) and the suffix, differing from the former in that it contri- 
butes nothing to the meaning of the word, and from the latter in that it is not 
used in the formation of groups of words (cf. τρέμειν, τρέσσε, trepidus, where 
“, 0, p, belong neither to the root nor to the suftix). 


352 A GRAMMAR OF NEW TESTAMENT GREEK. [ὃ 141-142 


moum (see Boisacq 57). For vowel gradation see Hirt Hdb. 89. On the 
interesting Biblical history of the word see Hort Comm. on 1 Pet. p. 77. 
Πόλεμος. Uncertain derivation. See Boisacq s.vv. πελεμίζω, πάλλω. 
With these should also come : 
Ποταμός (Lesh. πόταμος) 1.1}. *petd-, to move rapidly, fly, tumble (cf. 
ἐπτάμηνῚ, cf. also I.E. *pet- *pete-, Skt. pat-man, πέτομαι (ἐπτόμην). 


2. Nouns in -μή. 


This formation is closely parallel to that im -μός, and was 
used for verbal abstract nouns, many of which became concrete. 


(a) The verbal root is evident in γνώμη, δέσμη (for accent and meaning 
see Vocab. s.v.), δραχμή (--δράσσομαι, but for other theory see Vocab. s.v.), 
δυσμή (<dvvo. First=dvors setting, then=west), μνήμη, ῥύμη (-- ῥύομαι. 
For its Hellenistic meaning=street, see Rutherford NP 487-8), στιγμή 
(-:-στίζω <*orvy-4o), τιμή (--τίω, Tivo), φήμη (-- φημί). 

(b) The suffix -σμή is half concealed in αἰχμή (whence αἰχμάλωτος), 
which comes from *aiksma, and ὀσμή (cf. ὀδμή, p. 112) from ὄξω. 


(c) Other nouns are Ἰδοκιμή (see Vocab. s.v.), ζύμη (« ἔζῦσμᾶ or 
(ipa, LE. *ja(s)-mda, “juice,” “ pottage,”’ see Boisacq, s.v.), κἄλἄμη (see 
under 1 (f)), ὁρμή (<I.E. *ser-), πυγμή (see Boisacq, s.v. πύξ), which seems 
to be concrete (=fist) in the difficult passage Mk 7°. The verbal force 
is prominent in Barn 12? ( fighting, fight). 

Θέρμη (=Oépp-n) is not in place in this group. 


3. Adjectives in -pos. 
§ 142. These are mainly denominatives and compounds. 


(a) ΤὙπέρακμος (for derivation see p. 326 above). For meaning, how- 
ever, of this dz. Ney. see J. Weiss ap. 1 Co 1738, where the comparative force 
of the adjective is taken to be modal rather than temporal. So D. Smith 
(L. and L. of St. Paul, 268 n. 6) favours “‘ exceedingly lusty’ rather than 
κε past the flower of youth,” and cites tmepaxpatw=excel in youthful vigour 
(Athen. 657 Ὁ). ᾿Απόδημος, Τδίδραχμος, ἄζυμος, πρόθυμος, ἄμωμος 
(see Vocab. s.v.), ἡδύοσμος, ἄσημος (also εὔ-σ“, ἐπί- σ΄, παρά-σ΄“), δίστομος, 
tovvowpos, ἄτιμος (also βαρύ-τ', ἔν-τ΄, Τἰσό-τ'΄, πολύ-τ.), βλάσφημος, 
and εὔφημος, εὐώνυμος, and ψευδώνυμος. 


(Ὁ) Δίδυμος (Boisacq derives from ἔδι-, οἵ, δίς ‘‘ deux fois” +dv-pos 
“ apparenté de fagon peu claire ἃ duo”; cf. ἀμφίδυμος “ double.” Brug- 
mann /F xi. 283 n. suggests influence of νήδυμος), ἔρημος (<I.E. eré= 
separate, cf. dpaws, scarce, Lat. rarus), Τἤρεμος (1 Ti 22, see LS, and 
for vernac. use see Vocab.). The derivation of ἕτοιμος is obscure. (See 
under (4) below.) 


§ 142-143] SUFFLXES. 353 


(c) T’A@eapos (2 Pet 27 3:7, see Mayor én loc. and Vocab.) «-- θεσμός for 
Laconian θεθμός. For analogical invasion of o (from -σμός group) into 
conglutinates in -6-j.0s see Brugmann Grd. τι. i. 252. 

(d) Ἕβδομος represents a distinct suffix -μο-, cf. Lat. septimus, and see 
Brugmann (rd.? τι. i. 225 f. 


4. Adjectives in -(σ)ιμος. 

Those in -ἰμὸς begin with adjectives in which -yos was added to 
an i-stem; thus πρώϊμος (: πρωΐ), ὄψιμος (: *dyi) were formed from 
adverbs (for mpdipos, the correct rdg. in Jas 5’, see Ὁ. 73; Ruth. NP 
124); noun stems are recognisable in the Homeric κάλλι-μος and 
φαίδιμος (where -ἰ- represents an r- suffix d@aidp-ds), and thus the ending 
τίμος came to be detached and given as a new suffix. 

Examples in the NT are δόκιμος and ἀδόκιμος, σπόριμος, φρόνιμος, 
ὠφέλιμος. We may compare the proper name Τρόφιμος. 

The same suffix -yos joined to ἐΐ- stems, 1.6. to verbal abstracts in 
-σις, produced the suffix -σιμος, which has become very productive in 
late and MGr. (Brugmann-Thumb Gr. 219). For papyri, see Mayser 
Gr. i. 449. 

Thus βρώσιμος, θανάσιμος, χρήσιμος. Cf. proper name Ὀνήσιμος 
(with the play upon words in Philem 11). 

Notre.—1l. Debrunner Worth. 155 suggests that from -ἰ stems the 
suffix passed naturally to the verbal abstracts in -ovs in which the Attic 
effacement of the c- character (declension -ews, -εἰ etc.) favoured a further 
transference to other verbal nouns (6.9. μάχιμος : μάχη), and then to 
nouns other than verbai (e.g. νόμιμος (-ws): νόμος). (See also Solmsen 
Gr. Woréf. 49.) 

2. Ἕτοιμος (earlier ἑτοῖμος) might possibly come here, but Boisacq’s 
verdict seems against this (“‘ mot obscur; formation isolée”’). 


(2) -men-, -mon- (-μα, -μην, -μων). 
1. Neuters in -pa. 


§ 143. This very productive class of neuter nouns originates 
in I.E. -my, which is widely represented in many branches of 
the parent language. 

In Greek it appears as a dental stem, but the close connexion of -μα 
with -men- is seen both in comparison with Latin (e.g. ὄνο-μα, -paros, 
no-men, -minis) and in the continuance of the n- stem in verbal derivatives 
in -μαίνειν (see § 167) and in the cpds. in μων (see p. 355). : 


It is found with : 
(a) Primary verbal stems : “Appa, βῆμα, βλέμμα, ἐπί-βλημα, βρῶμα, 
γράμμα, δεῖγμα, δέρμα, διάδημα, ὑπόδ᾽, δόγμα, Ἰδόμα, Τέἔν-δύμα, δῶμα, 


354 A GRAMMAR OF NEW TESTAMENT GREEK. [§ 143 


θαῦμα, Τἀνά-θεμα, ἀνάτθημα, θρέμμα, ἴαμα, κάλυμμα, καῦμα, κέρμα, 
κλέμμα, ἔγκλημα, κλῆμα; κλίμα, ie κομμα, κρίμα, κτῆμα, κῦμα, 
λεῖμμα, ES μίγμα, μνῆμα, ὄμμα, ees leslie πνεῦμα, 
Ἱπόμα, πρᾶγμα, πτύσμα, πτῶμα, ῥῆγμα, ῥῆμα, σπέρμα, dae διά- 
στημα, στίγμα, σχῆμα, τάγμα, Τέν-ταλμα, τραῦμα, τρῆμα, σύν-τριμμα, 
Τἔκττρωμα (Rutherford NP 288f.), χάσμα, χρῆμα, ἵπερί-ψημα 
(§ 124 (b)). 

Norr.—In Hellenistic a great preference is shown for the short 
penultimate in such words as κλίμα, κρίμα, πόμα, which had already 
come in as doublets to «Aiwa, κρῖμα, πῶμα, under the influence of the 
feminine abstracts in -ovs (Sdua: δόσις ete.) and the verbal adjectives in’ 
-ros. See (@éros) above, p. 57, also Brugmann 67. 222. The two 
forms ἀνάθημα (votive offering) and ἀνάθεμα (curse) were both preserved 
for the convenient distn. of meanings: see Proleg. 46, Vocab. s.vv. 

(Ὁ) Denominative verbal stems: ’Ew-dyyeApa, αἴνιγμα, αἴτημα, 


U > 


Ταἰτίωμα (for class. airiaua, see Vocab. s.v.), Tédioynpa, ἀντάλλαγμα, 
ἁμάρτημα, Τἄντλημα, βδέλυγμα, βούλημα, γένημα, γέννημα, Ta-yvonpa, 
ἀ-δίκημα, δικαίωμα, δώρημα, Τὲἑδραίωμα, ἕλιγμα, Τἐξ-έραμα, Τέἐνέργημα 
(see Capes Ach. L. p. 348), ῥᾳδιούργημα, ἐπ-ερώτημα, ζήτημα; 
Τἥττημα (see Lightfoot, Lietzmann, J. Weiss ap. 1 Co 6’, and Vocab. 
8.v.), Τθέλημα, θυμίαμα, Τἱεράτευμα, Trrept-Kdbappa, Τόλο-καύτωμα, 
καύχημα, κήρυγμα, μίσθωμα, νόημα, νόσημα, οἴκημα, ὁμοίωμα, ὅρᾶμα, 
δι-όρθωμα (Ac 345 ΚΑΒ 33), κατεόρθωμα (ib. w), ὅρμημα, ἵμεσουράνημα, 
ὀφείλημα, ὀχύρωμα (Xen.), πάθημα, ἵπερίσσευμα, πλήρωμα, ποίημα, 
πολίτευμα, ἀ-σθένημα, σκήνωμα, στερέωμα, στράτευμα, τρύπημα, 
ὑστέρημα, ὕψωμα, φίλημα, φρόνημα, φύραμα, χάραγμα. 

(c) The suffix takes the form -σμα with -ἰζω, -ἀζω 
and other dental stems: from -it# come βάπτισμα, 
κτίσμα, νόμισμα, ῥάπισμα, Toxiopa, Τχάρισμα ; 

from -ἀζω, Τἀπαύγασμα, σέβασμα, σκέπασμα, Τἀπο-σκίασμα, 


“σμα 


φάντασμα, Τχόρτασμα ; 
from other dental stems, πλάσμα (πλάσσω-“πλαθιω)ὴ, ψεύσμα, 


(ψεύδομαι). 
Thence the suffix spreads (as with -σμός, p. 351) to nasal stems and stems 


of other verbs with perf. pass. in -σμαι: πτύσμα (πτύω), χάσμα (χάσκω 
or xaive), χρίσμα (χρίω), κέλευσμα (κελεύω), κλάσμα (κλάω), κύλισμα 
(κυλίνδω : “pa, 2 Pet 2% ΚΑΚΙ;Ρ), μίασμα (μιαίνω), Τκαταπέτασμα 
(καταπετάννυμι) ; also μέθυσμα (Herm. Mand. viii. 3: μεθύσκω). 


(4) A few words of uncertain derivation remain : 

Aipa<*aihua, *aioua, cf. Skt. is- isds, ‘‘ juice,” “ drink,” “ power.” 

Ὄνομα cf, Skt. nd@ma, Lat. nomen; Brugm. Grd.? τι. i. 234; Boisacq, 
8.U, 


Στόμα. See Grd.? τ, 383; Boisacq, s.v. 


§ 143] SUFFLXES. 355 


Σῶμα probably <*tuémnt, “swelling.” I.E. *tevd- which is seen ‘in 
σῶος (*7Fw-Fos) σάος (tFa-Fo-s) etc. Boisacq, s.v. 

Xeipa (χείμαρρος), see χειμών. 

Notr.—1. Although the close verbal connexion of these nouns in 
-pa is evident throughout the history of the language, in its earlier stages 
there was at the same time a tendency to use these formations for all 
kinds of nomina actionis, and another to lose the verbal conception in 
the most concrete of nouns. In the classical, and still more in the Hellen- 
istic period, a differentiation of meanings was observed in the use of the 
several formations: -ovs then expressed the verbal abstract (cf. Latin 
-- 0), -μός generally indicated the state, and - μα the result of the action (see, 
however, pp. 350, 374). Debrunner (Wortb. 157) points out other factors 
which played their part. Thus phonetic sympathies gave a preference 
for -ἡμα over -ημός, whilst -ἰσμός was preferred to -.cpa. The NT list 
shows 35 nouns in -μα as against one only in -μος, δῆμος, which goes 
back to very early times (see above, ὃ 141 (f)), and 23 nouns in -ἰσμός 
against 8 in -ισμα. He also finds distinctions due to the period, dialect 
and style of Greek writers. 

2. Thumb (Hellen. 216) calls attention to the productivity of this 
suffix in the Κοινή, and attributes it to Ionic influence. Where Attic 
uses ἀπόκρισις, ἔγκλισις, νίκη, νόσος, the Kown prefers ἀπόκριμα, ἔγκλιμα, 
νίκημα, νόσημα. He also remarks that outside the Κοινή this formation 
is most common in poetry, and specially in the Tragedians. 

3. See Capes Ach. L. p. 247, for relation between σύστασις in Polyb. 
ii. 42. 1 and σύστημα ib. 41. 15. 

4, See Helbing Gr. 113 ff. for LX X nouns in -pa. To his list Wacker- 
nagel ThLZ xxxiii. 641 adds ἀνάστεμα, ἀνταπόδομα, ἀφαίρεμα, ἀπόδομα, 
διάταγμα, δόμα, εὕρεμα, ἕψεμα, ζέμα, κάθεμα, παράθεμα, σύστεμα. See also 
Thackeray Gr. 80. 

5. For words in -μα in papyri, see Mayser Gir. i. 433 ff., to which 
add, 6.9., ἀγώνισμα C P Herm 121°, κατόρθωμα 125 iit, τέλεσμα 1271, 
δαπάνημα BGU iv. 112618, ὕδρευμα BGU iv. 1130”. 


2. Masculines in -μὴην and -μων. 


These suffixes were specially used in the formation of nomina 
agentis and adjectives. 
Aipyy, ποιμήν, Τἀρχι-ποιμήν. 
Τ᾿ Αρτέμων (: ἀρτάω), χειμών (-- χεῖμα), δαίμων (<*d*i, cf. δαίομαι), 
ἡγεμών ( : ἡγέομαι). Φιλήμων. 
Adjectives: ᾿Ελεήμων, οἰκτίρμων, ἐπιστήμων, ἀσχήμων, εὐσχήμων 
(3 σχῆμα). 
(9) -ηνῖ-. 
Α 51η8}} group of nouns may be mentioned here. 
pean Θέμις (> ἀθέμιτος) cf. Av. dami- (Brugmann Gr,* 
219, Grd.? τι. i, 254), and probably δύναμις. 


306 A GRAMMAR OF NEW TESTAMENT GREEK. [ὃ 144-145 


(4) Other Suffiwes in -n-. 
§ 144. -en, -on. 


For these nouns with their various ablaut grades, see above, 
pp. 134-6, also Brugmann Gr.* 219 f. 

It has been observed (J. Wright Comp. Gr. Gr. 
130) that this suffix was specially used in the forma- 
tion of nouns denoting (1) animate objects and (2) parts of the body. 

Under these headings come : 

(1) ᾿Αρήν (see p. 135), ἄρσην ; γείτων, κύων, τέκτων, τρυγών. To this 
group δράκων, λέων, θεράπων originally belonged, as the feminine forms 
-awa (=-av-1a) show. 

(2) The only NT words which represent this class are φρήν, σιαγών. 
Perhaps βραχίων may come in here. 

(3) A very old use of the suffix -ων (gen. -wvos) was to represent the 
possession of a quality or characteristic signified by the primitive noun. 
Examples of this usage are not here in evidence, but an extension of it 
is seen in the formation of surnames, nicknames, and, indirectly, in the 
patronymics in -ίων. 

We may quote Τίμων, Σίμων, Ἡρῳδίων. For other names in -or, 
see above, p. 146. 

(4) There is uncertainty about the origin of place-names in -ών (gen. 
-@vos) and words indicating locality. 

To this class belong ἀμπελών, Τἀφεδρών, κοιτών, μυλών (Mt 24* 
D et al.), Tvupdav, muddy. For ᾿Ελαιών see above, pp. 151 f., for 
Κεδρών (so in Josephus) see above, p. 149. 

(5) For the comparative suffix see above, pp. 165 ff. 

(6) For derivation of αἰών see Vocab. 16}. 


-nv and -wv 


(5) -no-, -na-. 
l. -vos, -v7. 
Ξ δ 146. As primary suffix im nouns and 
- τν os 
5. adjectives. 
(a) Nouns, e.g. αἶνος, ἀμνός, θρόνος, καπνός, οἶνος, πόρνος, ὕπνος, 
χρόνος ; ζώνη, κλίνη, πλάνη, σκηνή, φάτνη, φωνή ; δεῖπνον, κρίνον, τέκνον. 
(6) Adjectives, esp. verbal adjectives, eg. ἁγνός, δεινός, πτηνός, 
σεμνός, γυμνός, πυκνός. 
-s-no- is disguised in λύχνος (- ἔλυκσνο-ς), τέχνη (<*rexova). 
Ἴχνος (<iksmno-, see Brugm. Grd.2 1. i. 265, 245) belongs to the 
stems in -σ, see below, ὃ 158. 
-va is found in a few back-formations from verbs. 
Thus μέριμνα from μεριμνάω ; οἵ. ἔρευνα from “dw, γέννα 
from yevvaw. So Solmsen, Wortf. 39f., 238, following Wackernagel 
KZ xxx, 300, 314. 


“va 


§ 145] SUFFIXES. 357 


2. -avos, -avn, -avov is used in the formation of a number 


of nouns and adjectives. 
αντος, τῇ; τον : 
(α) κλίβανος, οὐρανός, στέφανος ; βοτάνη. δαπάνη, 


σαργάνη ; δρέπανον, λάχανον, πήγανον, φρύγανον. 

(b) ἱκανός, ὀρφανός. 

Norr.—According to Solmsen Wortf. 257f., ἔχιδνα is a substantiv- 
ised adj. in -dvos (cf. Maxedvds=Makedavos), Ἐἐχίδνη becoming ἔχιδνα 
under the influence of Spaxawa. 


3. -elvos arises from -vos added as a secondary suffix 
to neuter stems in -es (-εινός---εσ-νο-ς). 


, 
-εινός : 5 ss 5 
Thus ἐλεεινός : ἔλεος, ὀρεινός : ὄρος. 


In φωτεινός the same suffix has been transferred to the stem of the 
Attic φῶς (φωτ-)} from φαεινός (<*paFeovos): φάος. 

Analogy played a larger part in the formation of σκοτεινός either 
in the direct influence of φωτεινός or through the influence of the neut. 
φῶς in changing ὁ σκότος to τὸ σκότος in the early classical period. See, 
however, Brugmann’s theory (p. 126 above). 

Metewds (<méroua) and ταπεινός (for deriv. see Boisacq) are 
probably analogical formations. 

Notrt.—l. As *eo-vo-s has produced this group, so *ao-vo-s is not 
altogether unfruitful, as evidenced by σελήνη (Attic for Lesb. σελάν νᾶ), 
which is the substantivised fem. of ἔσελ-ασ-νο-ς : σέλας. 

2. WH App.” 161, “‘ Adjectives that in the best MSS have -ἰνός for 


, > , , eer Jt} 
-εἰνὸς ae ορινος, TKOTLVOS, φωτινός. 


4. -ovn is a suffix used in words to denote tools, instru- 
ments etc., as βελόνη (perhaps also ὀθόνη), 
and also (with a different accent) to form 
abstracts, 6.0. ἡδονή. 


, 


The derivation of ὀθόνη is uncertain. Some trace it to *FeAovn 
(accounting for its form by vocalic assimilation). Cf. Zd. fra-vadamna, 
Fr. vétue, A.S. woéd. Others find evidence of Semitic origin, and cf. Hebr. 
POX yarn. See Boisacq 687, 1119. Its use by Luke, as also that of 


βελόνη, is remarked by Hobart, pp. 218, 61. 


5. -uvos, -vvn. 


Originally an extension of the -vo- suffix 
to u- stems, this never became productive 
in the formation of nouns, 


-uvos, τύνη 


358 A GRAMMAR OF NEW TESTAMENT GREEK.  [§ 145-146 


Of the four which occur in the NT, αἰσχύνη is a back-formation from 
αἰσχύνω which was formed from the much older αἶσχος. 

Βόθῦνος (</bhed(h), bhod(h)) is a rare and late equivalent for 
βόθρος and is probably influenced by the cognate βαθύς. 

In κίνδῦνος and ὀδύνη the -"- has not been explained, but for the 
latter see the suggestion in § 111. 


6. -cuvos, -σύνη. 


The adjectival suffix -συνος, of obscure 
origin, was never fruitful except in the forma- 
tion of abstract nouns in -σύνη. 


-cuvos, σύνη 


Starting from such adjectival formations as δουλόσυνος, an extended 
form of δοῦλος, the suffix -civn was added to nouns ending in -os, to 
nouns and adjs. in -ης, to adjectives in τῶν and even to nouns in -ἰς, which 
were all treated as o- stems. 

In the NT we have : 

Δικαιοσύνη ( : δίκαιος), Τἐλεημοσύνη ( : ἐλεήμων). 

᾿Αφροσύνη, εὐφροσύνη, σωφροσύνη, Τταπεινο-φροσύνη ( : “φρων). 

᾿Ασχημοσύνη, εὐσχημοσύνη ( : “-σχήμων). 

Ἱερωσύνη (--ἱερεωσύνη Attic -ἱερηξ-οσύνη) : ἱερεύς, 
became in Hellenistic a model from which -ωσύνη was 
conveniently applied to avoid a series of short vowels, e.g. Τἀγαθωσύνη, 
Τἁγιωσύνη, Tpeyakwourn. 

For frequency in Hellenistic see Vocab. s.v. ἀγαθωσύνη, and for NT 
see SH Romans, p. 404. 


-ωσύνη 


7. Temporal Adjectives in -wos. 


§ 146. These can be traced to locatives or adverbs 
in -7 (see Brugmann Grd.? τι. i. 270). Thus ἐαρινός : 
ἔαρι, πρωϊνός : πρωί. In the class. age -ινός was applied to other stems, 
ἑσπερινός (which appears in Xen.) superseding the earlier ἑσπέριος 
formed regularly from ἕσπερος. The detached suffix was then used 
analogically for all kinds of time descriptions; e.g. Τόρθρινός ( : ὄρθρος) 
Lk 24” for earlier form ὄρθριος (ib. ἘΣ et al.), also Herm. Sim. v. 1" (see 
Lobeck Phryn. 51), and ταχινός ( : τάχα). 

ἸΚαθημερινός (<xal? ἡμέραν, ὃ 109; see p. 158) survives in MGr 
καθημερνός. 

Φθινοπωρινός, see above, ὃ 106, and Mayor on Jude 13. 

Cf. νυκτερινός, quoted from papp. in Vocab. 4328 ; also Clem. ad Cor. 20°, 
καιροὶ ἐαρινοὶ καὶ θερινοὶ καὶ μεταπωρινοὶ καὶ χειμερινοὶ ev εἰρήνῃ μεταπαρα- 


διδόασιν ἀλλήλοις, 


, 
“ινος 


§ 146] SUFFIXES. 359 


8. Adjectives of material in -ἰνος. 


A large group of adjectives signifying material, 
origin or kind was formed with the suffix -wos, in 
prehistoric times originating with 7- stems, but from Homer onward found 
used with all kinds of stems. 

The NT supplies ἀκάνθινος, Τἀμαράντινος, ἀνθρώπινος, βύσσινος, 
δερμάτινος, ἐλεφάντινος, Τθύϊνος, Τκόκκινος, κρίθινος, λίθινος, Τμύλινος, 
ξύλινος, ὀστράκινος, πύρινος, σάρκινος, τρίχινος, ὑακίνθινος, ὑάλινος. 

To these may be added σάρδινος (Rev 45 Ῥ et al. for σάρδιον) and 
Ἰσμαράγδινος, the adjective in each case being used with λίθος (under- 
stood). 

Two oxytones obviously belong to this rather than the preceding 
class : 

᾿Αληθινός the only adjectival derivative in this group, from ἀληθής, 

Πεδινός (πεδίον) which probably follows the accent of its antonym 
opewvos (see 3. above). 

The activity of this suffix in late Greek (cf. οὐθαμινός P Flor ii. 170? 
(iii/A.D.) ; ὑαμινός, (see Herwerden Lez. s.v.) suggests a possible explanation 
of the form συκάμινος (see above, p. 153). 


=LVOS 


9. -ivos, -avos, -nvos. 


Ethnica and adjectives signifying ‘ belonging 
to,” which have the suffix -ivos, may have originated, 
as Debrunner (Wortb. 162) following Κα. ΒΙ. ii. 296 suggests, in words 
formed from the names of Greek towns in S. Italy and Sicily, and were 
therefore of Latin origin (e.g. ᾿Ακραγαντῖνος, Ταραντῖνος). 

In NT we have ᾿Αλεξανδρῖνος (which, however, in the best MSS is 
᾿Αλεξανδρινός), ᾿Εφεσῖνος (296 and another cursive ap. Rev 2'), and the 
unquestionably Latin Λιβερτῖνος. 

-avos and -ηνός, which were not native to Greece, 
came to be used from the time of Alexander as 
ethnica, for Asiatic towns and districts. Thus ᾿Αδραμύττειον : ᾿Αδρα- 
μυττηνός (WH Αδραμυντηνός), ᾿Ασία : ᾿Ασιανός, Γερασά : Γερασηνός, 
Δαμασκός : Δαμασκηνός, Ναΐζαρά: Ναξαρηνός. Also Μαγδαλά : Μαγ- 
δαληνή. Used as nouns for place-names, ᾿Αβειληνή, Μελιτήνη (Ac 28! B), 
Μιτυλήνη. 


κεῖνος 


-ανός, -ηνός 


10. -cavos. 


Latin is also responsible for the extended suffix 
-avés. Stems in -ὖ took the normal Latin ter- 
mination -anus to denote a follower of a party leader. The partisans 
of C. Marius were thus termed Mari-ani, those of Cn. Pompeius Pompet- 


“τανός 


360 A GRAMMAR OF NEW TESTAMENT GREEK. [ὃ 146-147 


ani. It was an easy transition to Casar-iani. Thus Χριστιανός, 
Ἡρῳδιανός. 

Notr.—R. 8. Radford, ‘“‘ The Suffixes -Anus and -inus”’ in Gilder- 
sleeve Studies, 1903, pp. 95 ff., shows (after Schnorr v. Carolsfeld, 
Archiv f. lat. Lex. i. 177-194) that the formation was so regularly from 
the nomen in -ius (or -ia stems, cf. Fimbrianus) that “from the second 
half of the Augustan period on new formations are made only in -ianus.”’ 
On p. 98 he catalogues “ extension of the purely Roman suffix -anus 
(-idnus) to other than Roman words.” Thus Herculanus, ef. Plaut- 
aneus, Trophonianus (Cic.), Thyonianus (Cat.), Hannibalianus, Hasdru- 
balianus and many others. The suffix when added to names of persons 
makes a possessive, closely allied to gen. for cases of adoption, but more 
restricted than corresponding gen. and tending to recur in set phrases. 
He cft. “ the Clayton-Bulwer treaty,” etc. [Germ. -sche]. Varro says 
(LL 9. 42, 71 Sp.) from Faustus should be Faustinus, Faustianus could 
only rightly be from Faustius. But Cato says Quirinianus. Cicero 
avoids these new forms in the orations and philosophical works, and 
uses them twice only in his letters. From -6n- base he has Pésontanus 
and Neronianus in or. and phil. The suffix -inus was practically obsolete 
in the silver age, The conclusion is “ that the possessive adjectives in 
-anus are later formations of the language, and that the gentile adjective 
(Titius, Aelius) originally possessed a much broader, perhaps a wholly 
unrestricted, use. This older usage was always retained in the ritualistic 
language, in the poets and in the authors of archaic tendencies.” 4 


C. Liquid Suffixes. 


(1) l- Suffizes. 
1. -lo-, -la- 

§ 147. The simple suffix -λος, -λη was not fruitful, 
though a number of words survive. 

Nouns: Αὐλός, ἄθλος (d-Feb-dos) (:- ἀθλέω), ζῆλος, ξύλον, ὅπλον 
(<LE. *sop-lo-m), σπήλαιον (cf. Hom. σπέος), σπίλος, στύλος (Skt. 
sthurd-s). 

᾿ομίχλη, φυλή. 

Adjectives: Βέβηλος, δειλός, δῆλος (-- ἔδειηλος), δοῦλος, στρεβλός 
(:-. στρεβλόω), τυφλός, φαῦλος. 

NotEe.—BeBn)\os, an old t.t. of religion, from ν' βη- in βαίνω, means 
accessible and so profane, as opposed to ἄβατος, inaccessible and so 
sacred. Σπίλος Boisacq (p. 693) connects with κηλίς, Lat. caligo, 
squdlus, <1.E. *sq*d, or else from LE. *spéi- *spt-. Rutherford NP 87, 
identifies with σπιλάς [<I.E. *spéi-, *spt=to be pointed, cf. Lat. spina, 
spica, pinna (*pitnd), Boisacq, p. 896 1.1, and “‘ tracks σπίλος rock through 


dos, τ-τλη 


1 Mr. E. E. Genner calls attention to such formations as ᾿Αττικιανός from 
** Atticus.” 


§ 147] - SUFFIXES. 361 


an easy gradation of meanings historically consecutive from the beginning 
to the close of Greek literature. Originally meaning rock, it came to 
signify successively porous rock, rotten-stone, clay, and clay-stain, till Paul 
could employ it metaphorically as in Eph 5??, and Dion. H. apply it to 
men with the meaning dregs of humanity.” 


2. -aXo-, -ελο-, -ἰλο-. 


These are more numerous. 

Nouns : Αἰγιαλός, διδάσκαλος, ὕαλος (see Thumb 
in DAC i, 553b): κύμβαλον, πηδάλιον, Τσκάνδαλον, 
Τσκύβαλον (<I.E. *squb-, see Boisacq s.v., but see § 105 for derivation, 
which removes it into class of cpds.): ἀγκάλη (<dyxos), κεφαλή 
(cf. Gotha gibla, OHG gebal), κρεπάλη, σπατάλη (>toTratahdaw), φιάλη. 

Adjectives : ᾿ἀπαλός (cf. Lat. sapere, and ὀπός juice), μεγάλ-η. 

"Ayyehos, ἄμπελος, μυελός ; ἀγέλη, νεφέλη (Lat. 
nebula) ; εὐτράπελος (:- λία, § 106). 

᾿Ὀργίλος, ποικίλος : κοιλία (--κοῖλος <*xdFidos, ef. 
Lat. cavus). 

Nore.—Atyahds. Hirt IdgF (1917) xxxvii. 229f., starting from 
Hesych. αἶγες ᾿ τὰ κύματα Δωριεῖς (also Artem. Oneirocrit. ii, 12), leaps 
to aif§=shore. Then in such a line as Hom. Jl. iv. 422 ἐν αἰγιαλῷ may 
originally have been ἐν αἰγὶ ἁλός, 1.6. “* on the shore of the salt-flood.” 
If this association was frequent, they flowed together into one word, at 
least in pronunciation. We should then have a connexion which was 
transformed into αἰγιαλῷ, to form again a nom. from this, αἰγιαλός. 

_ Σκάνδαλον, Moulton Exp 1' xxvi. 331, connects with Skt. skand 
“leap,” “‘ spirt,” Lat. scando, O.Ir. scendim “1 spring,” and holds that 
σκάνδαλον existed before σκανδάληθρον (Arist. Acharn. 687), though not 
occurring in literature. 

Κρεπάλη (for spelling see p. 81) possibly connected with κραιπνός 
swift, impetuous. So Boisacq 506, who, however, rejects the theory of 
a primitive ἔκραπνιος, related to καρπάλιμος, and is unconvinced by the 
etymology *xpa- head +- πάλη. 


-ahos 


-ελος 


τιλος. 


3. -ἅλό-, -ηλό-, -ωλό-. 


These productive adjectival suffixes, in which -a\os came to be 
specially appended to stems with corresponding verbs in -av, and -ηλός 
to stems with similar verbs in -eiv, are not represented in N.T. 

The later use of -7Aos to form nomina agentis has 
provided κάπηλος (whence “λεύω), τράχηλος (properly 
“carrier,” so Brugmann Gr.* 231. For another derivation see Boisacq, 
s.v.). The detached suffix is appended without any semantic reason in 
ὑψηλός. 


-ωλός -ωλύς is represented in ἁμαρτωλός and εἴδωλον. 


πηλὸς 


362 A GRAMMAR OF NEW TESTAMENT GREEK. [§ 147-148 


4, -υ(λ)λο-. 


The diminutive conception which was always 
closely related to the -/o- suffixes and is specially 
evident in the mass of Latin diminutives in -wlus, -ellus, -illus, comes out 
in Greek, specially with words in -vAos and -vAdos. In proper names this 
suffix generally marks either a pet name or an abbreviation of a compound 
appellative. 

Δάκτυλος (<*dar-x-vdo- <*dnt-go-, cf. MHG zint, Germ. Zinke, Zacke), 
acc. to Brugmann (JF xi. 284 ff., αν." 159, Grd.? τι. i. 484), originally meant 
“little tooth.” 

Σταφυλή (<I.E. *stmbi-), see Boisacq 90, 91, 903. 

{BypudXos (Rev 2150, so also in Tob 13!7 B). For the form βηρύλλιον 
(Ex 28°) see above, p. 343. 

Τέρτυλλος =Lat. Tertullus, dim. of Tertius. 

Τρωγύλιον. For this “΄ Western and Syrian” reading in Ac 201° see 
WH App.? 98. For spelling see WS 47, and Ramsay CRE 155. 


-uhos 


5. -αλέος. 
λέ It is curious that this extended suffix which was 
rece widely used in the Κοινή (Br.-Thumb Gr. 231) does 
not appear in NT except in inferior MSS for νηφάλιος (see p. 76). 


6. -αλις. 


This seems to be a feminine suffix in names of 
animals, in δάμαλις (perh. through masc. δαμάλης, 
where V is dap- (cf. Ir. dam<Celt. *damo-s), πάρδαλις (masc. πάρδος 
<I.E. *prda, cf. Skt. pfdaku-h). So also potxadis (LXX, Test. XII 
Patr., Plut.; for Attic μοιχάς, see Lobeck Phryn. 452). 

Quite distinct is σεμίδαλις, cf. Lat. simila (*sem-). Possibly from I.E. 
sem- <*bhs-em- (: ψωμός), cf. *bhas- to pound, which fits the meaning 
fine wheat. This rare word (Rev 181%) in addition to citations in Gr.-Th. 
is found Justin M. Ap. i. 37 and papp. (see Vocab. s.v.). 


-αλις 


(2) r- Suffixes. 


§ 148. The most important words with r- suffixes are the 
nomina agentis in -τὴρ and -twp (see §§ 150, 152) and the 
related formations in -tpos, -tpov (see ὃ 152), which are dealt 
with below. 


1. -ro-,. -ra. 

This suffix was partly connected with the -p in 
the nom. and ace. sing. of neuters, 6.9. ἄνυδρος : ὕδωρ, 
ἄλευρον : *a\cFap, and with the suffix -er, 6.9. αὔρα : ἀήρ. 


=pos, -pa 


§ 148] SUFFIXES. 363 


Nouns : ᾿Αγρός, ἄγρα, ἄκρον, αὔρα, ἀφρός, δῶρον, ἕδρα, (ἔνεδρα, see 
above, p. 125), ἔχθρα, κλῆρος, κόπριον, κοπρία, λεπρός, λέπρα, μῆρος, 
νεφρός, ὄμβρος, πέτρα, πήρα, πλευρά, πρῷρα, σταυρός (Brugm. Grd.* τι. 
i. 351), τέφρον (τεφρόω), χώρα. 

Adjectives: Αἰσχρός, ἄνυδρος, ἄφεδρος (:- Τἀφεδρών, see § 115), 
ἐλαφρός, ἐρυθρός, ἐχθρός, λαμπρός, μακρός, μικρός, νεκρός, νωθρός 
(see Vocab. s.v.), ξηρός, πενιχρός (see Vocab. s.v.), πηρός (ἀνάπηρος), 
πικρός, σαπρός, σκληρός, σύνεδρος (: συνέδριον), ὕγρός, χλωρός, ψυχρός. 


2. -ἄρο-, -ερο-, -opo-, -υρο- (principally with adjectives). 
See Brugmann Gird.* τι. i. 347 f. 

isos, Pant Κιθάρα; ἵλαρός, καθαρός, λιπαρός, ῥυπαρός, 

χλιαρός. 

Ἑσπέρα, ἡμέρα ; πενθερός, πενθερά ; βλαβεροός, 
ἱερός, καρτερός (whence ‘pyats), φανερός, φοβερός. 

With different accentuation, ἐλεύθερος (ἐ- prothetic vowel, Brugmann 
Gr.* 173, cf. Lat. liber. Boisacq 242, derives from *edevdos, generation, 
people, cf. O. Slav. ljudije). 

BépBopos (2 Pet 255, a classical word found else- 
where in bibl. Gr. only in Jer (LXX) 458): λοίδορος. 

For the -Fopos cpds. (e.g. θυρωρός) see ὃ 105. Θεωρός (>°péw) belongs 
probably to that group ; see below, p. 391. 

"Axupov ; ἄργυρος (Skt. drjuna-s, bright, white), 
ὀχυρός (ὀχύρωμα), ψίθυρος (Ἱψιθυρισμός). 


τερος, -ερὰ 


τορος 


εὔρος 


3. -Gpo-, -npo-, -ὔρο-. 

As with -lo- suffixes, combination with stems ending in long vowels 
(e.g. ὀδυνηρός : ὀδυνάω, ὀδύνη) formed a class to which adjectives derived 
from other stems were added by false analogy. 

-apos Φλύαρος (φλύω). 

Αὐστηρός (avo, dry up, <*avoyw, Lat. haurio, haus- 
tum), αὐχμηρός (αὐχμός, drought), ὀκνηρός (: ὀκνέω, 
ὄκνος), πονηρός (: πονέομαι, πόνος). Τολμηρός (>ToApnpotépws) formed 
regularly from τολμάω. The origin of the noun σίδηρος is unknown. 

"Aykupa (ἀγκών), κολλύρα (>keAAUpLOV, see p. 78) : 
ἰσχυρός (ἰσχύ:). 


τῆρος 


-Upos, -upa 


4, -np, -wp. 
For ἀήρ see Brugm. Grd.? τι. i. 339, ἀνήρ, ib. 332 f. 
, +Katnywp is a Hellenistic back-formation from 
ἘΠῚ ΜΝ κατήγορος (see Thumb Hellen. 126, and Brugmann- 
Thumb Gr. 210 n.’, also p. 127 above). 
VOL. II. PART III.—25 


364 A GRAMMAR OF NEW TESTAMENT GREEK. [§ 148-150 


5. For -repos, see below, ὃ 153, p. 369. 
6. -dhro-, -dhra-, and -dhlo-, -dhla-. 


This suffix seems to be related on one side to the -7o-, -lo- suffixes, and 
on the other to those in -tro-, -tlo- (see ὃ 152 below). Brugmann (Grd.? τι. 
i. 377) regards this -dh- (as in -Θθμός, ὃ 141 above) as probably a root deter- 
minative. ne 

Ὄλεθρος (ὄλλῦμι, «ὀλ-νῦ-μι), ὄρθρος, σκυθρός 
(:-. σκυθρωπός) «ἔσκυσθρός (by dissimilation) : σκύ- 
Copa, φόβηθρον (Lk 211: BDW, see pp. 110, 369), 
κολυμβήθρα (: κολυμβάω). 

There is no NT example οὗ -θλο-, but γενέθλιον (Mart. Polyc. 183) is a 
derivative of γένεθλον and appears in disguise at Mk 6?! D (see p. 112). 


-9pos, -θρον, 
-Opa 


D. Suffixes with Labial Stops. 


1. -τ-. 
att § 149. Apart from ποταπὸς (see Prol. 95 and pp. 112, 
271 above), stems in -7- are κώνωψ, μώλωψ, ὕδρωψ 
(ὑδρωπικός) (genitives in -wzos), σκόλοψ (gen. -οπος), λαῖλαψ (gen. 
-avos). (See further Sturtevant, Cl. Phil. vii. 425 ff.) 

As ἄνθρωπος has not been dealt with among the cpds., it may be 
mentioned here. Boisacq rejects the old derivation ἀνήρ -ἰ- ὧψ (which 
leaves the -θ- unexplained), preferring <*avdp-hwro-s “having the 
appearance of a man,” where the second element is related to the Gothic 
sailvan, to see, Lat. signum<*seq"no-m. 


2. -bho-, -bha-. 

ΤῸΝ Ἔριφος is the sole representative in the NT of the 
old class of animal names in -qos (cf. ἔλαφος). 

We may note three other words. Ἔδαφος <*é5-, ν 864), by the law 
which forbids successive aspirates under the influence of ἔδος has passed 
from the masc. type to the neut. with stem in ἐδαφε(σ)- (Brugm. Gird.? 
Ir. i. 390). Κύλαφος is preserved in Τκολαφίζω. Ψῆφος pebble, polished 
by the churning of the waves, suggests connexion with I.E. *bhs*-bh- 
>Skt. bhas-, “to pound,” “crush.” The same root appears in ψάμμος 
(=sand). See Boisacq, s.vv. 


E. Suffixes with Dental Stops, 
(1) The -t- Suffixes. 
1. Nomina agentis in -τήρ, -Twp, -τῆης. 


§ 150. The original distinction between these suffixes was 
that simple verbs formed nomina agentis in -τήρ or -τωρ, and 


§ 150] SUFFIXES. 365 


noun compounds in -τῆς. Compound verbs followed the usage 
of the simplex when the preverb was felt to be an integral part 
of the verb; otherwise they took -rns. Lonic-Attic very early 
substituted -τῆς for -rnp, with the result that in Hellenistic 
-τῆς is left in possession. 


A few of those in -r7p, -rwp, mostly religious and 
legal terms, survive: σωτήρ, φωστήρ, φυλακτήρ (pre- 
served in φυλακτήριον). 

ἹΚοσμοκράτωρ, ὑπαντοκράτωρ (see ὃ 107), +kTHTwP, 
πράκτωρ, ῥήτωρ. Σπεκουλάτωρ is a Latin loan word. 
Οἰκήτωρ occurs in Clem. ad Cor. 144. Οἰκητήριον is witness for the 
earlier -τήρ. ᾿Αλέκτωρ (superseded in Att. by ἀλεκτρύων, but reappearing 
in Kown: see Rutherford NP 307f. and Vocab. s.v.) should come here, 
if Kretschmer (KZ xxxiii. 560) is right in deriving it from ἀλέξω. See 
Boisacq 43. 


Notr.—For difference of accent in -ryp and -rwp see Brugmann 
Grd.? τι. 1. 331. 


-τήρ 


πτωρ 


The vast majority of these nomina agentis are 
formed in -tns. 

From verbs in -dw : ᾿Ακροατής, κυβερνήτης, πλανήτης, τολμητής. 

From verbs in -έω: Αὐλητής, ἐπιθυμητής, Τκαθηγητής, Τκατα- 
φρονητής, μετρητής, μιμητής, ποιητής, Τπροσαίτης (a late word— 
Plut., Lucian—by haplology for προσαιτήτης, see Hirt Handb. 172),. 
Ἱπροσκυνητής (NT ἁ.λ. Jn 4%. See Vocab. s.v. and LAH} 99f. (2101), 
for probable pre-Christian use), Ἰσυνζητητής (NT 4.A. 1 Co 130, elsewhere 
only in the citation in Ignat. Hph. 181). 

From verbs in -όω: Ζηλωτής, Τλυτρώτης (LXX, Philo, Just. M., 
Act. Thom.), ᾿τελειωτής (ἅ.λ. in Heb 122—apparently coined by the 
author). 

From verbs in -evw: Βουλευτής, ἑρμηνευτής, Τὀλοθρευτής (dA. in 
1 Co 1019 ; for form see p. 71), παιδευτής. 

From verbs in -df#: ΤΒιαστής (NT ἁ.λ. in Mt 1122. A late form for 
Bards, common in Pindar. Wetst. cites from Eustath. and Aretaeus 
alone. Gr.-Th. quotes Philo Agric. 19, where Cohn and Wendland print 
divisim, Bias τῶν), δικαστής, Ἰστασιαστής (for class. στασιωτής). 

From verbs in -ἰζω: ᾿Ανδραποδιστής (see p. 286 for derivation of 
this class. word; in NT only 1 Ti 110), ᾿βαπτιστής (NT and Joseph. ; see 
Vocab. 102 δ), βασανιστής, Τδανιστής (for form, see p. 77), ᾿ Ἑλληνιστής, 
Τέξορκιστής (apart from NT ἅ.λ. Ac 191, only found in Joseph., Lucian, 
and eccles. writers), Τεὐαγγελιστής (NT and eccles., but see Vocab. 
s.v. for inscr. possibly non-Christian), θεριστής, Tkeppatiotys (see N. 1 
below), tktlotns (NT ἅ.λ. 1 Pet 41%; see Vocab. s.v. for use in Κοινή), 
λῃστής, μεριστής (NT aA, Lk 12, To Pollux and eccles. add Κοινή 


-τῆης, τής 


366 A GRAMMAR OF NEW TESTAMENT GREEK. [§ 150-151 


exx. in Vocab. s.v.), toadmorys (for class. σαλπιγκτής, Lob. Phryn. 191), 
ὑβριστής, ψιθυριστής. 

From verb in -v¢@: ΤΓογγυστής (NT ad. Jude 15; elsewhere only 
LXX, Sym., Theod.). 

From other verbal stems : Γγώστης (in which o is inserted before τ 
as in δυνάστης and ψεύστης), διώκτης (NT ἅ.λ. 1 Ti 113, elsewhere 
Did. 5°, Barn. 207, Sym. But the LXX cpd. ἐργοδιώκτης is ‘‘ profane,” 
as shown by a iii/B.c. papyrus: see Vocab. 166a), Τδότης (d.A. 2 Co 9? 
(LXX), for class. δοτήρ ; see Vocab, s.v. for pre-Christian inser. ἐκδότης), 
Τἐμπαίκτης (only LXX semel, 2 Pet 3%, Jude 18: see Mayor Commentary, 
in loc.), ἐπενδύτης (see n. 2 below), ἐφευρετής, κλέπτης, κριτής, μαθητής 
(cf. με-μάθ-η- κα), νομοθέτης (a T.P. cpd., see ὃ 106), ὀφειλέτης, παραβάτης, 
Ἰπλήκτης, ἱπροσωπολήμπτης (ὃ 105), προφήτης, ὑπηρέτης (§ 129 (b)). 

ΝΟΤΕ.---Ἰ. Κερματιστής (peculiar to NT and related writings) is 
formed from *i¢@, a classical verb for to cut small, which gained the mean- 
ing in Hellenistic to coin inio small money (Anth.), to exchange (for papp. 
see Vocab.) ; Τκολλυβιστής, a late word (Rutherford NP 499, common in 
papp., see Vocab.), has no corresponding verb, and must owe its form to 
analogy, probably with κερματιστής. The noun κόλλυβος is of Semitic 
origin. 

2. As with nouns in -τήρ, we can trace an easy transition from nomen 
agentis to instrumental meaning in some words, e.g. ἐπενδύτης, μετρητής. 


2. Derivatives in -rns from nouns. 


δ 151. The denominatives include ναύτης, πρεσβύτης. The suffix 
is found after a variety of stems, some of which have given rise to complex 
suffixes which were freely used in the production of analogy formations. 
Thus : 

-érns : from stems in -o-, οἰκέτης (οἰκότης only found in vulgar 
curse tablet, Meisterhans* 117), ὑσυμφυλέτης (for Attic φυλέτης, 
from φῦλον or φυλή). So εὐεργέτης (see below, under cpds.). 

τώτης: from -o- stems, δεσμώτης, ἰδιώτης: from -a stems, 
στρατιώτης, Τσυνηλικιώτης (for class. ἡλικιώτης). 

τίτης : from -ἰ- stems, πολίτης. By analogy μεσίτης and ἃ 
group of appellatives which τεχνίτης and τραπεζίτης (WH 
-εἰτης) represent in the NT. These two influences may account 
for ᾿Αρεοπαγίτης (ὃ 106). The common ethnica in -irys are 
naturally reinforced by the Semitic ethnica in » (fem. ΠΥ 


though the spelling -είτης is better supported (WH App.? 161). 
See the list above, p. 150. 

Νικολαΐτης, a follower of Nixé\aos—an easy extension of the 
idea of “* membership” that constitutes the preceding classes. Μαργαρίτης 
may (through its association with precious stones) be quoted as the one 
NT example of a class of nouns with this suffix used of kinds of stone (e.g. 
mupitns, αἱματίτης ; cf, Lat. anthracites (Pliny)). 


§ 151-152] SUFFIXES. 367 


3. Compounds im -της. 


The large class with -στάτης as second member of the compound 
is represented by ἐπιστάτης and πρωτοστάτης (a K.D. cpd., ὃ 106). 
So αὐτόπτης (a K.D. ορά. § 106) and ἐπόπτης (ὃ 120) represent another 
group. Three T.P. cpds. (ὃ 105) are ἀρσενοκοίτης (from ἄρσην, κοιτῆ), 
οἰνοπότης, Τφρεναπάτης. Probably to the same class belongs χρεοφει- 
λέτης (for form see p. 73, also Lob. Phryn. 691). 

Notr.—The accentual variations within these groups call for some 
notice. Brugmann (JF ix. (1898) 368 n.*) draws attention to the corre- 
spondence between the masculines κλέπτης, γυμνήτης etc., with their pen- 
ultimate accent, and the abstracts βλάστη, ἀήτη etc., just as κριτῆς 
etc. correspond to the abstracts ἀκτή, μελετή etc. But the connexion 
is not shown. The fullest treatment is in Fraenkel, Geschichte der 
griechischen Nomina agentis, ii. 199-215. Debrunner Worth. 175 f., while 
suspecting that the ancient grammarians added to the confusion by 
unreliable transmission in individual words as well as groups, offers the 
following general rules. Derivatives in -τῆς from nouns are barytone, 
also the old nomina agentis in -rns, and the compounds, together with a 
few uncompounded formations as ἱκέτης and κλέπτης. Oxytone are 
those formations in -rjs which have taken the place of an earlier -rnp. 
Thus οἰκέτης, πολίτης, δεσμώτης, ἐπόπτης, οἰνοπότης; but βουλευτής. 
Other factors, however, may break through this partition. Thus the 
Hellenistic δότης (for Homeric δοτήρ) is under the influence of -δότης, so 
often the final member in cpds. (e.g. μισθαποδότης). On the contrary, 
κριτής (also ὑποκριτής), Which replaced Dor. κριτήρ, maintained a sturdy 
independence of δικαιο-κρίτης etc. because of its legal use. It will be 
seen that generally those formed directly from a verb are oxytone. 


4. Abstract nouns in -της. 


§ 152. These are nearly all formed from 
adjectives in -o-, with the result that the 
suffix -ότης is regularly attached to the con- 
sonantal stems as well. 


Abstracts 
in -τῆς 


From nouns in -o- : ΤΘεότης, Τἀδελφότης (note extension of meaning 
to conerete and collective sense. See Vocab.). 

From adjectives in -o-: Τ᾽ Αγιότης, Τἁγνότης, ἁδρότης, αἰσχρότης, 
Κγυμνότης, Τἀδηλότης, Τθειότης, ἱκανότης, Τἱλαρότης, ἰσότης, καθαρότης, 
καινότης, Τκυριότης, λαμπρότης, [ματαιότης (see Vocab.), ᾿μεγαλειότης, 
νεότης, ὁμοιότης, ὁσιότης, παλαιότης, σεμνότης, σκληρότης, τελειότης, 
Ττιμιότης, χρηστότης : ἁπλότης (from -dos). 

Τ᾿ Αφελότης (from -ς, for which see ὃ 107) appears in the Κοινή together 
with class. ἀφέλεια, e.g. Vett. Val. p. 24015. (See Vocab. s.v.) 

Πιότης (from πίων), Τένότης (from εἷς gen. ἑνός). 


368 A GRAMMAR OF NEW TESTAMENT GREEK. _[§ 152 


From adjectives in -ύς : Βραδυτής, εὐθύτης, πραὕτης. 

Notr.—For the accent in βραδυτής see Br.-Th. Gr. 180 and Debr. 
Worth. 184 π.1. The distinction seems to rest upon an I.E. principle of 
accentuation. In Att. and Hellen. there was a tendency to assimilation, 
e.g. Att. κουφοτής : tpaxurns. Hellen. γλυκύτης ete. : νεότης. 


5. Other nouns in -rnp. 


-τὴρ These fall into three groups : 


(a) Closely connected with nomina agentis, already 
discussed under 1, are a number of terms for instruments, tools, etc. 
(ef. words in -rpov (under 6. below). 

ΤΝιπτήρ.----ΤῊϊ5 word for basin, found first in Jn 135, and elsewhere 
only in writings influenced by the Johannine narrative of the foot-wash- 
ing, is formed from the late verb virrw (class. νίζω) : cf. κρατήρ, κρητήρ fr. 
κεράννυμι. Στατήρ (Skt. sthdtar, Lat. stator, I.E. *stha-, cf. ν sta- of torn), 
used first of a weight, then of a coin. The connexion with nomina agentis 
is here traceable. 

Χαρακτήρ (χαράσσω), originally a tool for engraving, then of the 
die or mould, then of the stamp or impress, as on a coin or seal; so Heb 15. 


(Ὁ) Terms of relationship: Θυγάτηρ, μήτηρ, πατήρ. On the “ re- 
lative ’” force of -rnp cf. under 7. below. 

(c) A few other nouns of various meaning : 

᾿Αστήρ (Skt. stdr-, Lat. stella <*stér-la. The a- in Gr. and Arm. astl 
is not prothetic. I.E. *astér, gen. strés. Boisacq, s.v.). 

Taotip.—Brugmann (JF xi. 272 n.) suggests connexion with yév-ro 
“he held,” yéuo “I am full of anything,” γέμος, youos etc. For the 
o he compares Skt. vasti-, vanisthi-, Lat. vé(n)sica; Germ. wanst, Lat. 
venter. ; 

Notre.—Brugmann (Gr.* 180) suspects that the recessive accent in 
μήτηρ, θυγάτηρ, for *untnp, *Ovyarnp (Skt. mata, duhitd), is due to the 
influence of the vocative μῆτερ, θύγατερ, in the same way that personal 
names prefer the recessive accent on that account. (See p. 59 above.) 


6. Nouns in -Tpos, -Tpa, -Tpov. 


=Tpos, =Tpa, These are closely related to the -ter-, -tor- 
~Tpov suffix of the nomina agentis. 
(a) -τρος, ἃ most unproductive suffix, is represented in NT only 
by tatpds ( : iarnp, Ion. inrnp). 
(Ὁ) -rpa is seen in μήτρα, womb, where the instrumental force is 
evident. 
(c) -rpov is common in names for implements. Debrunner (Wortb. 


176 1.) traces three stages in use of this suffix: (a) instrument, (8) 
locality, (y) payment. 


§ 152-153] -  SUFFIXES. ' 369 


Under (a) we have ἄγκιστρον (Mt 1737, Ignat. Magn. 111), ἄροτρον, 
ἀμφίβληστρον (ὃ 112, 1 (b)), ἔσοπτρον (ὃ 117), κέντρον, λουτρόν, μέτρον, 
φόβητρον (Lk 211, So most MSS, see above, pp. 110, 864. Both 
suffixes are old, but this spelling was preferred because of dissimilation of 
aspirates. BI-D ὃ 35, 2). Under (8) θέατρον (so Ac 199-%1; then 
spectacle as in 1 Co 4°). Under (y) λύτρον and ἀντίλυτρον. 

Notr.—l. The accent of λουτρόν, like that of another Homeric word 
δαιτρόν, distinguishes it from almost every other noun in -τρον. (For 
the group of cognates, λουτρών and λουτήρ, with which it is so closely 
allied in meaning, see J. A. Robinson ap. Eph 5?°.) 

2. The papyri show that -rpov was still active. Thus for (a) 
ἐγκοίμητρον ““ counterpane,’ P Petr ii. 92°; for (y) the very common 
φόρετρον “ cost of carriage,” e.g. P Rylii. 209° (ii/a.D.). 


7. -τερος. 


§ 153. This was confined in the historical period 
to the formation of the comparative of adjectives, but 
the suffix -tero- was used at a much earlier stage to indicate ideas that were 
relative rather than absolute. The I.E. comparative in -tero- did not stand 
for an intensification of the positive, but represented a comparison. So 
the form in -tero- expressed no absolute property of a thing, but a 
relationship. Pairs of correlatives or opposites were thus formed, 


=TEPOS 


E.g. ἀριστερός---δεξιτερός (cf. Lat. sinister—dexter). 
ἡμέτερος- --ὑμέτερος (cf. Lat. noster—vester). 
épéarepos—ayporepos (“ἡ on the hills *»—“‘ on the plain ’’). 


Streitberg (JF xxxv. 196f.) contends that the suffix -ter- is to be 
explained in the same way. It is seen at once in the terms for family 
relationship πατήρ, μήτηρ, θυγάτηρ, φράτηρ, which all point to the charac- 
ter of one person in relation to another. We can understand why such 
nouns should be provided with the same suffix as the so-called compara- 
tive formations. It is but one step further to claim that the same holds 
good of the nomina agentis. These originally were concerned with ideas 
which had no unqualified but only a relative meaning, and would be 
unthinkable except in some external connexion. Just as the term 
‘* father ” involves a relationship to another person, “ child,” so “‘ giver” 
(datar-, Sornp, Serwp) implies relationship to a second party. Thus the 
suffix -ter- is appropriate here also. 

Comparatives. See above, pp. 165-7. The ordinal δεύτερος is of the 
same formation. 

Pronouns: ᾿Αμφότερος, ἕτερος, see p. 182 above. Tldrepos, which of 
two, survives in NT only in adverbial form, πότερον, whether, Jn 1717, 

The adj. ἀλλότριος owes the secondary suffix -ἰο- to its original 
-antonym ἴδιος. The primary suffix -rp- is the weak grade of -rep-. For 
similar ablaut changes see Brugmann Gir.* 228, Gird.? τι. i. 165, 329. 


370 A GRAMMAR OF NEW TESTAMENT GREEK. [§ 154 


8. Nouns and adjectives vn -τος, -τη, -τον. 

§ 154. These suffixes were chiefly used in the parent I.E. to form verbal 
adjectives and ordinal numerals. In Greek the verbal connexion is less 
strong than in some cognate languages, and -ros is not the mark of the 
perf. pass. partic. (cf. Lat. -tws), but is free from tense and voice con- 


nexions (Proleg. 221). The verbal attachment was weak enough to 
allow the suffix to be added to noun stems. 


(a) Verbal adjectwes. For formation see 
pp. 188, 224, and for accent, p. 224. 


=TOS 


These may be grouped conveniently under the conjugation classes 
(supra 184 f£., infra 381 ff.). Often the neg. adj. alone occurs. 


From verbs in I. a. (a): Τ᾿ Αδιάλειπτος (see Vocab. s.v.), Τἀκατάλυτος, 
Τἀκατάπαστος (p. 253, Vocab. s.v.), Τἀκατάσχετος, ἀκώλυτος (-ws, Vocab. 
8.0.), ἄμεμπτος, ἀμεταμέλητος, ἀνεκτός, Τἀνένδεκτος, ἄπιστος, Τἀπόδεκτος, 
ἄπταιστος (Xen. and later), γραπτός, δεκτός (Vocab. s.v.), ἐκλεκτός, 
Τΐεστός, Τθεόπνευστος (p. 255, Vocab. s.v.), twapetoaxtos (Vocab. s.v.), 
πιστός (Vocab. s.v.), Tmvixtés, Tmpoondutos (p. 237, also Vocab. s.v.), 
ἵσυνεκλεκτός. 

(b) ᾿Απρόσιτος. 


I. β. (b) ᾿Ακατάστατος (Hippocr. and Hellen. writers), ἀμετάθετος, 
ἀνεύθετος, ἀσύνετος, ἐγκάθετος, ἔκδοτος, ἔκθετος, εὔθετος (§ 106), 
μετάδοτος (εὐμετ΄, ὃ 106), παράδοτος ({πατρο-“, § 105), περίστατος (εὐπ΄, 
§ 106), συνετός. 

II. a. (a) ᾿Αναμάρτητος, ἀνεπίλημπτος, Τἀπερίτμητος, ποτός (>noun, 
πότος, p. 254). 

(b) Δυνατός, whence ἀδύνατος (ὃ 106). 

II. 8. (a) ᾿Αμετακίνητος, πρόσφατος (ὃ 125). 

(Ὁ) “Akpatos, ἄρρωστος (Hippocr. and Hellen.), ἄσβεστος, 
στρωτός (λιθό-“). 

Ill. Τ᾽ Ἀπερίσπαστος (7-ως). 

IV. (a) ᾿Αρεστός, εὐάρεστος (Vocab. s.v.), -βρωτός (σητό-, Τσκωληκό-), 
παθητός. Μεθυστός:- ἀμέθυστος (in Plut.=“‘ not drunken,” then used for 
a remedy against drunkenness, and so of the stone (ἡ ἀμέθ.)). 

(Ὁ) “Ayvworos (ὃ 106), Τἀκατάγνωστος, γνωστός, διδακτός. 


(c) Αἰχμάλωτος (-.αἰχμή and ἁλίσκομαι, a T.P. cpd., ὃ 105), θνητός. 
There are no verbals from verbs in Classes V. and VI. Those in the 
next Class follow the subdivision of verbs in yo : ye given below, § 160. 


VII. From verbs in -dw. ᾿Αγαπητός, Τἀρτιγέννητος, ἀνεξεραύνητος, 
ἀμώμητος, ἀόρατος (ὃ 106), γεννητός, Spatds. From χρήομαι, 
χρηστός (ρ. 265). 


§ 154] SUFFIXES. 371 


in -€o. ΤΑγενεαλόγητος, Τἀλάλητος, Τἀμετανόητος, Τἀναπολό- 
γητος, ἀναρίθμητος, ἀνέγκλητος, Τἀνεκδιήγητος, Τ-ἀνεκλά- 
λητος, ἀνόητος, ἀόρατος (ὃ 106), ἀρκετός, αὐθαίρετος, Τάχειρο- 
ποίητος, ϑυσνόητος, Ἐπαίνετος, ἐπιπόθητος, Τεὐλόγητος, 
κλητός, παράκλητος, ᾿ποταμοφόρητος, (for papp. see Vocab. 
8.0.), στυγητός, χειροποίητος. 
in -όω. Μισθωτός. 
in -ύω. -θυτός (Τεἰδωλόθυτος, ἱερόθυτος), -puros (ἔμφυτος, 
νεόφυτος). 
in -εὐω. ᾿Απαίδευτος, ἀσάλευτος, ᾿δυσερμήνευτος, ἵλαξευτός, 
σιτευτός. 
ἴπ -αἰνω. ΤΑμάραντος, ἀμίαντος, ἀπαράβατος (Ruth. NP 
367. For meaning, see Westcott ap. Heb 7*4 and Vocab. s.v.), 
ἀπέραντος, ἄφαντος (see Vocab. s.v.), ὑφαντός. 
in -ive (-n- yo). ᾿Αδιάκριτος, Τἀκατάκριτος, ἀνυπόκριτος, ᾿Ασύν- 
κριτος, Ταὐτοκατάκριτος. 
ἴῃ -ύνω. Τ᾽ Ανεπαίσχυντος. 
in -apo, -εἰρω. ᾿Ακάθαρτος, φθαρτός, ἄφθαρτος. (Here also 
may come ῥητός (=@s), ἄρρητος, ἀναντίρητος. See p. 235.) 
in-\Aw. ᾿Απόβλητος. 
in -πτω. fF Ayvapos (=dyvarros, see Vocab. s.v.), Τάκατα- 
κάλυπτος, ἄμεμπτος, ἄνιπτος, ἄραφος [ΕΞ πτος), κρυπτός. 
in -σσω. ΤΑνυπότακτος, ἄτακτος, βδελυκτός, πλαστός, 
τακτός. 
in -ἀζω. Τ᾽ Ἀνεξιχνίαστος, ἀπαρασκεύαστος, Τἀπείραστος, 
ἵδυσβάστακτος (for guttural form, see p. 230 and Vocab. 
pp. 106, 173), θαυμαστός, σεβαστός. 
in -ἰζω. ΤΑστήρικτος, ἀθέμιτος (for older ἀθέμιστος), TAOTHPLKTOS, 
ἀχάριστος, εὐχάριστος, Τσιτιστός ; also ἄσωτος (-ws, § 106). 
But ἀθέμιτος (for older ἀθέμιστος) from θέμις. Χριστός 
is from χρίω. 
Notr.—1. The general rule for accentuation is that verbal adjectives 
in -ros are oxytone when uncompounded, or when compounded with a 
preposition if they denote possibility (three terminations). Otherwise the 
compounds are proparoxytone (and of two terminations). But as the 
passive meaning can so easily pass over into that of possibility, we often 
find paroxytones with two terminations for oxytones with three termina- 
tions. For exceptions to these rules see K.BI. i. 538. Schmiedel (WS 69) 
explains ἐκλεκτός as derived direct from ἐκλέγω (with no corresponding 
simplex), so also εὐλογητός from εὐλογέω. This probably explains 
συνετός. Συνεκλεκτός is influenced by its simplex above. For ἀπόδεκτος 
and ἐκλεκτός, swpra, p. 58. 
2. ᾿Αμέθυστος and διάλεκτος preserve their adjectival force, as is shown 
by their feminine gender, due to the supply of λίθος and γλῶσσα respectively. 


372 A GRAMMAR OF NEW TESTAMENT GREEK. [§ 154 


(b) Nouns, abstract and concrete—The masc., fem., and neut. 
of these verbal adjectives often came to be used as abstract 
nouns. Many, however, have the concrete meaning. 


a. Θάνατος, Kovioptos, κοπετόξ, μαστός (see p. 110 above, νότος 
(see Boisacq, s.v. vaw), πλοῦτος, πότος, πυρετός, ὑετός, φόρτος, 
χόρτος. Στρατός (-“στόρνυμι) is represented in NT by compounds (e.g. 
στρατηγός, στρατολογέω). Σῖτος (<*Wiros : Wiw=pound. So Prellwitz. 
Boisacq gives this with alternative etymologies), ἐνιαυτός (<Hom. 
eviavw=rest in. Hence original meaning was solstice, i.e. place of rest in 


the sun’s progress. See Boisacq s.v., and Brugmann JF xy. 89 ff., xvii. 
319 f.). 


B. ᾿Αρετή, βροντή (βρέμω), γενετή, κοίτη, κρύπτη, 
τελευτή, μελετή (Barn. 1011). ‘Eopty possibly belongs 
to this group if the derivation <*Fe-Fop-ra is right. (See Boisacq s.v.) 


-τή 


y. Ἑρπετόν (ἕρπω), πρόβατον (προβαίνω), λεπτόν 
(λέπω) ; ἄριστον rightly comes in here, <*dpi<*a[cJepe 
(a locative form=in the morning) and ἕέστόν <éSo (=eat). See Brugm. 
KVG@ 458, and, for loss of « by ablaut, ib. 143. Ποτόν drink occurs in 
Did. 10°, Ignat. Tal. 23, Ep. Diogn. 6°. 


, 
=TOV 


(c) Suffic -wros—A special variety is the suffix -ωτός, 
originally belonging to the verbal adjective closely associated 
with verbs in -όω (see IV. (c) above), and then attached to 
other stems, even to nouns, in the sense of “ supplied with.” 


Λιβανωτός (<)i8avos) may come under this 
heading, if in Rev 8* it=censer, a view which Charles 
supports (ICC in loc.). See, however, Vocab. s.v. Κιβωτός is a word of 
Semitic origin and uncertain etymology. 

In later Greek the fem. form -o77 is fairly common as an elliptical 


-ωτός 


subst. with the meainng “τη846 of” or ‘“‘ coming from.” Thus μηλωτή 
sc. Sopa) coat of sheepskin («μῆλον sheep). See Mayser Gr. i. 454, 
Jannaris Gr, 297. 


(4) Superlatwes and Ordinals.—For the -tos suffix in the 
formation of superlatives see above, §§ 67-69, and for the closely 
related ordinals see § 72. 


The suffix -fo- appears chiefly in ordinals, e.g. ἕκτος (Skt. Sasthah) 
Lat. sextus). From ἔνατος, δέκατος etc., -aros became detached as a 
distinctive suffix and is found in two superlatives, πρῶτος and ἔσχατος 
(see p. 167). How the once productive superl. suffix -raros arose is not 
clear. Brugmann (Gr. 202) suggested φέρτατος -εφέριστος as the 
explanation, but this is dropped in ed.‘ (p. 238). Others account for it by 


§ 154-155] SUFFIXES. 373 


comparing the Hom. ὕστατος with Skt. wétamds and assuming that -rapo- 
became -raro- under the influence of -ros in -ἰστος. So Hirt Handb. 294. 
The suffix -.cros goes back to I.E. (cf. Skt. -istha-, Goth. -ista-), and is 
composed of -is-, weak grade of the suffix -zes- (cf. Lat. magis), and the 
-to- already mentioned. 


9. -τῷ, -τεί. 


Under this heading we find in NT only the group of adverbs in -vori, 
for which see p. 163. 


10. Nomina actionis in -ti-. 


§ 155. A very productive suffix from the 
I.E. period in the formation of primary verbal 
abstracts (nomina actionis) of the feminine gender. 


-σις 


For the change οὗ τ to o before « and ε at a primitive stage of the 
language, see Brugmann Gr.‘ 118. 

This suffix was originally added to the weak grade of the roots or 
bases (as with the verbal adjectives in -ros). In time, however, with 
the gradual passing of the ablaut distinction in many verbs, and partly 
under the influence of other forms of the verb, and of other verbal nouns, 
the suffix was attached also to the strong grade of ablaut. Thus βάσις, 
δόσις, but in Hellenistic ῥεῦσις came in for the earlier ῥύσις. It is there- 
fore worthy of notice that ῥύσις is the form in the NT. 


(1) Formed by adding suffix directly to root or base : 

Αἵρεσις (dv-, Si-, καθ-), βάσις (ἔκ-, κατά-, παρά-), βρῶσις, γένεσις, 
γνῶσις (ἀνά-, διά-, Τέπί-, Τπρό-), δόσις (ἀνταπό-, παρά-), δύσις 
(<dvvw=set), ἔγερσις, Τἔλεγξις, Τέλευσις, ἕξις, θλίψις, ἴασις, καῦσις, 
κλῆσις (παρά-, πρόσ-), κρίσις (ἀνά-, ἀπό-, διά-, Τκατά-, ὑπό-), κτίσις, 
λῆμψις (ἀνά-, ἀντί-, μετά-, πρόσ-), λύσις (ἀνά-, ἐπί-), ὅρασις, ὄρεξις, 
ὄψις, πόσις, πρᾶξις, πτῶσις, ῥύσις, στάσις (ἀνά-, ἀποκατά-, ἔκ-, ἐξανά-, 
ἐπί-, Τέπισύ-, ὑπό-), τάξις, φάσις (πρό-), φύσις, χρῆσις (Τἀπό-). 

In the following words the suffix is added to the compound verbal 


stem (as was the case with most of the compounds just given in brackets). 
They are put in a separate list as the simplicia are not found in NT: 


Ὕπαρξις, ἀνάβλεψις, Τἀνά-, ἀπό-, ἔν- δειξις, Τἀπέκ-, ἔν- Suors, 
On’ 5 3) , , ᾿ , ΕἸ , 5 , > , 
(-- δύω), ἄν-, ἄφ-, πάρ-, σύν- εσις, Τκατάσχεσις, GVTL=, ἀπό-, ETL-, 
μετά-, ἵπερί-, πρό-, Τσυνκατά- θεσις (θέσις in Herm. Vis. iii. 135), 
ἄφιξις, ἀπόλαυσις, ἀνά-, ὑπό- μνησις, Τκατάνυξις, ἄνοιξις, ἀνά-, κατά- 

παυσις, ἔντευξις, Τἀνά-, Τπρόσ-, σύγ- χυσις. 
(2) This suffix was then attached to denominative verbal stems, 


especially when the existing noun from which the verb was derived did 
not convey the abstract verbal meaning. 


374 A GRAMMAR OF NEW TESTAMENT GREEK. [§ 155 


Thus from verbs in -a : 

Τ᾽ ἀγαλλίασις (see below, p. 385), ἀπ-, συν-, Τύπ- άντησις, γέννησις, 
Τἐνδώμησις (pp. 73, 307), Τκαύχησις, κοίμησις, κυβέρνησις. 

From verbs in -έω: 

᾿Αγανάκτησις, Τἀθέτησις, Τἄθλησις, Ταΐνεσις, δέησις, διήγησις, Texdi- 
κησις, ἐνθύμησις, Τἐπιπόθησις, Τέσθησις (but see p. 133), ζήτησις, 
κατοίκησις, κίνησις, Ἱπαρατήρησις, ποίησις, Τπροσκαρτέρησις, πτόησις, 
ἱσυμφώνησις, Τύστέρησις, φρόνησις. 

The short vowel in the penult of αἴνεσις, as the long vowel in θέλησις 
(--θέλω), is probably due to the influence of the aor. ἤνεσα, ἐθέλησα 
(see Brugm. G7.* 239). 

From verbs in -όω: 

Τ᾽ Ανακαίνωσις, BeBatwous, i βίωσις, δι-, ἐπαν- όρθωσις, δικαίωσις, 
Τἐκπλήρωσις, Τέἐρήμωσις, Τκατασκήνωσις, Τλύτρωσις, Τμόρφωσις, Τνέ- 
κρωσις, ὁμοίωσις, πύρωσις, πώρωσις, ταπείνωσις, τελείωσις, Τὑποτύπωσις, 
Τφανέρωσις, Tpuciwars. 

From other verbal stems : 

Κόλασις ( : -ἀζω), Τκατάρτισις (: -ἰζω), ἀποκάλυψις (: -πτω), αἴσθησις 
(: αἰσθ-άνο-μαι), αὔξησις (: αὐξ-άν-ω, fut. αὐξήσω), ἅλωσις (: ἁλίσκομαι, 
fut. ἁλώσο-μαι). συνείδησις (: σύνοιδα, fut. -εἰδήσω), Ὑπεποίθησις ( : πείθω, 
πέποιθα). 

Notes may be given on three words. 

“Aduats chain is of uncertain etymology, and may have no connexion 
with these nomina actionis. 

Πεποίθησις. ‘‘ Substantives in -ovs from the perfect stem were not 
used by Attic writers”? (Rutherford VP 355). 

Συνείδησις (see above ὃ 127). For history of this word see Norden 
Agnostos Theos, 136 n.'. It is noteworthy that with one doubtful excep- 
tion Epictetus prefers the participle τὸ συνειδός (cf. Bonhéffer, Epiktet τι. 
das NT’, 156). 


(3) The old form of the suffix -τὶς survives in a few old words only. 
Πίστις (-- πίθ-τις), νῆστις fasting, where νη is the result of contraction, 
high grade né coming before é in Anlaut, »/ ed-, see above, p. 287. 

The masc. μάντις (Herm. Mand. xi. 2) was originally fem. abstract 
(ef. Lat. hostis. Brugmann Gr.? 239). 

Notr.—1. -ovs nouns in Hellenistic no longer represent action merely, 
but also result. Thus κτίσις creation or creature, γνῶσις the result of 
insight, as well as insight itself. 


2. “*-ous is apparently concrete in αἴτησις ΟΡ Herm 73 118, but οἰκοδόμησις 
in 83° is nomen actionis; so πλάκωσις 9415, δήλωσις 10129. But κτῆσις 
1016, οἴκησις 119 itil’, σύμπτωσις often, κράτησις 119 iv?®, BeBaiwors 33}, 
μέτρησις v3, ἄθλησις verso iii!® [are concrete]. "Αξίοσις 1b.2° less clear. 
Πρᾶξις, legal execution (BGU iv. 1115*8 etc.). Thumb (Dial. 373) says 
-σις and -μα were signs of Lonic influence in Attic prose.’”’—-J. H. M. 


§ 156] SUFFIXES. 375 


(2) The -d- Suffixes. 


δ 156. Though originating in the I.E. period, this suffix 
had a specially extended use in Greek. 


1. Nouns in -αὃ-, -06-. 


1. As a root determinative it appears in such 
formations as παρα-στά-ς, -ddos, which became a 
model for formations in -ds, -ados. Thus ἱκμάς (Lk 85, see Vocab.), 
λαμπάς, σπιλάς (Jude 12: see Mayor in loc.), στιβάς (Mk 118, see Swete 
in loc. For στοιβάς and στυβάς see p. 76 above). Aopkds, a woman’s 
name, means “* gazelle,” and is an instance of a form modified by popular 
etymology, (opxas (=I1.E. *iork-s) being conformed to supposed connexion 
with δέρκομαι. 


-as (-ad-) 


2. Patronymics and place-names are represented by Ἡρῳδιάς, 
TiBepids, Τρῳάς, and Ἑλλάς. 

3. For the numerals puptds and χιλιάς, see above, pp. 169, 176. 

4. The old class to which ποδαπ΄ς, ἀλλοδαπός 
etc. belonged is represented in the NT by ποταπός, 
which has been modified by popular etymology under the influence of 
πότε. (See Proleg. 95.) 

5. The ethnica in -is are represented only by 

ge Ge) Ἑλληνίς (: masc. Ἕλλην), and “EBpats (a peculiar form 

for the more usual ‘ESpaixés, ἢ, 6v), which is found twice in the LXX 
(4 Mac 12? 1015, ἡ ‘E. φωνή) and in the NT in Acts ter (ἡ “EB. διάλεκτος). 

Σαμαρεῖτις in Joseph. =the region of Samaria, but in Jn 49 is the fem. 
of Σαμαρείτης. 


-ὅταπος 


6. Feminine appellatives in -ἰς are Τσυγγενίς (from m. -7s), Ἱμοιχαλίς 
(Hell. form of Att. μοιχάς, see above, ὃ 147, from m. μοιχός), Γπορφυρόπωλις 
(from m. -7s). 

7. A diminutive meaning may be traced in θυρίς ( : θύρα), κεφαλίς 
( : κεφαλή), πινακίς (Lk 168 C3D)). Cf. Plummer in loc: ‘ All four forms, 
πίναξ, πινακίς, πινάκιον, and πινακίδιον, are used of writing-tablets, and 
πινακίδα is v.l. here. But elsewhere in NT πίναξ is a “dish” or 
** platter.” ’ 

8. Apart from these groups we have a large number of nouns with 
the -ἰς termination. ᾿Ακρίς, ἀσπίς, ἀτμίς (cf. ἀτμός, «-ἄω--ἴο blow), 
Τβολίς (Heb 1230 (LXX) minusc. pauc.), ἐλπίς, κλεῖς, λεπίς, μερίς, παῖς 
(<mdFis), παγίς, ῥαφίς, ῥυτίς, σανίς, σφραγίς, σφυρίς (see above, p. 109). 

Ἴᾳασπις, a word of Pheenician origin (see Boisacq, and cf. Hebrew 
Maw). Ἶρις <*Fi-pi-s, I.E. *ui-ri-s. 

Napowis (see above, ὃ 123). The Atticists condemned the use of this 
word in the derivative sense (see NP 265). 


376 A GRAMMAR OF NEW TESTAMENT GREEK. [8 156-157 


Πατρίς, originally poet. fem. of πάτριος, then subst. (= πατρία yn), 
Σύρτις, Awis. 

9. -ris has become the regular feminine for nomina agentis in -rns, 
e.g. προφῆτις, προστάτις, and for denominatives in -rns, 6.0. πρεσβῦτις. 


2. Conglutinates with -a8-, -16-. 


(ὃ These are represented in the NT only by the 
eis neuter nouns in -ίδιον, for which see above, p. 346. 
3. Nouns in -δον-. 


The sole NT representative of this formation is 
-dwv (-δον-) : : Ὃ : ΤΕ 
χαλκηδών. Found in Rev 2119 alone in Biblical 
Greek, it is the name given to a copper silicate found in the mines near 
Chalcedon. The place name itself is a derivative of χαλκός copper. 
It is doubtful whether σινδών should come under this heading. It 
seems to be an Oriental loan word; cf. Hebr. 1 “linen wrapper.” 


4. Adjectives in -ὠδης. 

These have been given in § 107 above (p. 283) under compounds. 
The fondness of Hermas for words of this formation is striking : ἐρημώδης 
(Sim. ix. 261), κροκώδης (Sim. vi. 15), μαστώδης (Sim. ix. 1*), πυροειδὴς καὶ 
αἱματώδης (Vis. iv. 3°), κρημνώδης, ἀκανθώδης, τριβολώδης (Sim. vi. 2°), 
ἀκαν θώδης (Mand. xii. 13- * et al.). 

(3) The -0- Suffixes. 

See above, p. 364. 


F, Suffixes with Guttural Stops. 
(1) Lhe -k- Suffixes. 
1. Primary nouns in -κη. 


§ 157. A few words have this as a primary 
ae suffix, 6.9. θήκη, νίκη. 


νίκη <*ni- according to Osthoff MU iv. 223 4., who cft. Skt. ni-ca-h. 
But this is very doubtful ; see Boisacq 671. 


2. Adverbs in -ιξ, -a€. 


er: Πέριξ (§ 124 (c)) and ἅπαξ (p. 286) are 
ts nom. sing. of adjectives petrified as adverbs 
(cf. ἀναμίξ, πατάξ, Brugmann Gr. 207). 


§ 157] SUFFIXES. 377 


3. Nouns with stems in -ak-, -εκ-, -nk-, -UK-, -tK-. 


ΝΟΌΣ ites pce, “θύρας, κόραξ, τος φύλαξ, χάραξ : 
Ee crac ἢ -ex-: ἀλώπηξ ; in -ηκ- : σκώληξ ; in -υκ- : κῆρυξ 
(for accent, see above, p. 57); in -(α)ικ- : γυνή ; in -ἰκ- 
HALE, Φοῖνιξ, φοῖνιξ, χοῖνιξ. 

᾿ς The history of the relation between γυναικ- and γυνή is uncertain, 
but that it goes back to prim. I.E. is shown by Armen. kanai-. The 
accent in γυναικός may be derived from the monosyllabic ἔβναικ- 
(*Bvaxés). See Brugmann JF xxii. 171 ff., Gr.4 242. For etymology of 

ἀλώπηξ, see Brugmann Gird.? τι. 1. 474. 


4. Denominative adjectives and nouns in -xo-. 


Φυσικός (di-ci-s), GAuKds (GAs), μαλακός ; 
φάρμακον ; φυλακή. 

᾿Αλυκός is supplanted in Hellenistic by ἁλικός, really a distinct word. 
See above, p. 80, and Vocab. sv. (Apart from Jas 312 the classical 
form is found in the Bible only in the name for the Dead Sea, Num 3132, 
Deut 8317.) 

Pdppakov<*hapua<*bhrmen-. So Brugm. Grd.2 τι. i. 485. See, 
however, Thumb in Brugm. Gr.* 241 n.*, and Boisacq 1015 n.! 

Madakés<I.E.*m'l -κό-5, <mela’-, to grind, pulverise (see Boisacq, 
8.vv. μαλθακός, μύλλω, βλάβη). For the meaning of this word in 1 Co 69 
see Deissmann LAH 150 (7164) n.4, Vocab. s.v. 


“KOS, - κη, -κον 


5. Denominative adjectives in -caxos, formed from nouns in 
-to-, -ta, on the analogy of -ιάς : -is, -ιάδης 


, 
=LOKOS 
: -dOns, -ιάζειν : -ἔζειν. 


ἸΚυριακός (κύριος), Τοἰκιακός, are both Hellenistic formations, 

For the contemporary use of xvpiaxds=imperial, and for the origin 
of the use of ἡ κυριακή -- ΠΟΥ 8 day in Rev 119, see Deissmann BS 217 ff. 
LAE 362 ff. (2358 ff.), and Vocab. 364. 

Οἰκιακός (Mt 1025 86 only in Gr. Bible), a Κοινή formation found in 
papp. (see Vocab.) in the sense of a member of a household. Οἰκειακός, read 
v.22 by CDMU (v.** UP), is a late formation from οἰκεῖος, meaning 
his own. (Found in Plut. Cie. 20). 


6. Denominative adjectives in -ἰκός (after -vos the most 
productive of adjectival suffixes in Greek), 
from prim. I.E. -igo-, as in Skt. parydyikd-s 
(=strophic) from parydyd-s (=strophe), cf. Lat. modicus : 
modus. 


, 
=LKOS 


(a) In Homeric period these were mostly ethnica, which continue to 
be formed in this way. 


378 A GRAMMAR OF NEW TESTAMENT GREEK. [§ 157 


᾿Αχαϊκός, Γαλατικός, “EBpaixds, Ἑλληνικός, ᾿Ιουδαϊκός, ᾿ἸΙταλικός, 
Λευιτικός, Ποντικός, 'Ρωμαϊκός. 

φοινΐκη (p. 149) and Σαμοθράκη (-θράκη, BE) are of a different 
formation. 


(b) The extension of this suffix in the classical period probably arose 
with such words as φυσικός, μαντικός, where the suffix -xos was attached 
to an -ἰ- stem. 

The idea of ** belonging to” is seen in βασιλικός. (For the meaning 
in Jn 44° see Bauer in loc., in Jas 2° see Hort in loc. and Deissmann LAE 
367 η.3 (2362 n.°). For the stem see Fraenkel ZVS xlv. 222 f.). The suffix, 
which was a favourite with the Ionian sophists, came into common use 
with the Attic writers in the latter part of the fifth century B.c.,1 where 
the meaning of “ pertaining to,” “‘ with the characteristics of,’ became 
prominent. In the NT list we have Τἀρχιερατικός, Τἐθνικός, εἰρηνικός, 
ἱππικός, κεραμικός (see note 2 below), κοσμικός, ἵλειτουργικός (see 
LAE 170, 276), λογικός (see Vocab. s.v. For Rom 12! see Lietzmann HNT 
in loc.; for 1 Pet 25, Hort in loc.), μουσικός, tpudtkds, νομικός, Τόνικός 
(NT, papp., inscrr., see Vocab. s.v.), πατρικός, προβατικός, σιρικός (by 
vowel assimilation for onpixds, see Mayser Gr. i. 150, WH App. 158, 
above p. 72 ;—really an ethnic adj. from οἱ Σῆρες), Στοϊκός, τυπικός. 
(T-Gs), Ττυφωνικός, ὑδρωπικός, ἵχοϊκός. 

From compar. adjectives come ἀνωτερικός, Τνεωτερικός, and from 
adv. καθόλου, καθολικός (in the titles of Cath. Epp. in late MSS. See 
Mayor Comm. James ed.” cclix). 

Κοινωνικός, originally “ social,” later acquired the meaning ‘‘ ready 
to go shares ”’ (so 1 Ti 015), and thus approximates to a nuance found in 
some of the adjectives in -τικός (see below (c)). 

Σαρκικός, ψυχικός, πνευματικός, σωματικός form an important group. 

Norr.—1l. The distinction in meaning between adjectives in -ἰκός 


and those in -.vos is generally maintained, the former connoting . . . -like, 
and the latter made of . .. . It corresponds to that found in the English 
suffixes -y and -en: e.g. leathery, leathern, earthy, earthen. ‘* The ter- 


mination -ἰνος denotes a material relation, while -ἰκός denotes an ethical 
or dynamic relation, to the idea involved in the root” (Plummer ap. 
1 Co 3!, where capxivos is deliberately chosen in distinction from 
σαρκικοί in v.°). The true reading is preserved in NABC*D* 33; 
σαρκικοῖς (D3EFGLP) is an obvious correction. The same contrast 
with πνευματικός has led to the substitution of -ἰκός for -ἰνος in Rom 
714 (x°LPw); a similar misunderstanding is answerable for -ἰκῆς (>) 
in Heb 7!® On the other hand -ικός is right in Rom 15?’, 1 Co 3° (bis) 
(-ἰνοι D¥FG) 9:1, 2 Co 115 (-ivy FG) 101, 1 Pet 2%. In 1 Co 34 ncLP have 


1 Fraenkel, ut supra, 205 f., gives statistics showing the relative frequency 
in Euripides compared with Sophocles, in Thucydides : Herodotus, and in 
Isocrates : Iseus. It is specially common in Plato, Xenophon, and Aristotle, 
and in scientific terminology. 


§ 157] SUFFIXES. 379 


σαρκικοί, where ἄνθρωποι is the true reading. (See Westcott ap. Heb 715 
for true distribution of these words.) For confusion of meaning in these 
suffixes see Vocab. s.v, ξύλινος. 

2. Κεραμικός earthen (<xépapos clay) is to be distinguished from 
κεραμεικός of a potter (<xepapevs). It was a late form, deprecated by 
Phrynichus (see Lob. 147), for class. κεραμεοῦς, which was already 
undergoing change in Hellenistic (-μαῖος, Polyb., -μειος Plut.). Fraenkel 
shows (ib. 221) how closely κεραμεικός and -ἰκός approximated in 
meaning. In view of P Lond 1215% (=i. p. 112) ἀπὸ τρόχου [κε]ραμικοῦ 
and Ps 28 ὡς σκεῦος κεραμέως, Rev 257 (σκεύη κεραμικά) is probably an 
itacism (see above, 76 f.). 

3. The form ἀρχιερατικός follows the classical ἱερατικός, which is 
influenced by the verb ἱερατεύω, although there is no verbal stress in 
the adjective. 

4. Ἡλίκος, πηλίκος, τηλίκος (τηλικοῦτος) preserve a suffix -λι- (cf. 
Lat. qualis, talis) to which the secondary suffix -xos is attached. 


(c) From nomina agentis in -τῆς were 
formed many adjectives in -τικός, in which 
the verbal force was strongly present. These verbal deriva- 
tives took the same suffix with either an intransitive or a 
causative force. 


-=TLKOS 


Thus αἱρετικός capable of choosing (Plato) and so factious (Tit 3, 
where the current use of aipeois =secta, factio, has coloured the meaning, 
see Parry, Comm. in loc.), ᾿διδακτικός apt at teaching (1 Ti 3°, 2 Ti 2532), 
elsewhere only in Philo. The class. διδασκαλικός appears even in Vett. 
Val., and survives in MGr (see Vocab.). Κριτικός able to discern, 
Tmapadutikds the late and vernacular word (probably formed on the 
analogy of ἀναλυτικός, διαλυτικός, ἐκλυτικός, from Arist. onwards—for 
exx. see Fraenkel ib. 216) always used by Mt and Mk, whereas Lk retains 
the medical term παραλελυμένος, συστατικός («συνίστημι) post-classical 
in sense of ‘‘ constructive,’ and more often, as in 2 Co 3! a t.t. for 
commendatory (letter), Τπροφητικός (-«προφήτης), apart from Rom 16°, 
2 Pet 119 and Patrr., only in Philo and Lucian. 

Norg.—l. Βιωτικός (<fido) as first used by Aristotle =fit to 
live, lively, and shows analogy with most words in this group. But its 
regular use in Hellenistic (Polyb., Diod., Philo, Plut., Artem.), condemned 
by Phryn. (Rutherford NP 459), appears in Lk 21%, 1 Co 6°, where 
it is adj. corresponding to Bios (see Lightfoot Notes 211, Field Notes 
171). This meaning of ‘‘ worldly,” “‘ secular,” “* business,” “* everyday” 
can be illustrated from the unliterary as well as the literary Κοινή (papp.- 
Vett. Val. etc., see Vocab. s.v.) 

2. {Muotikds, that crux interpretum in Mk 14°, Jn 12%, if a Greek 
word, is either (a) from πιστός (<7eiOe) fit to be trusted, genuine, though 
elsewhere of persons, as in Artem. On. ii. 32, 66, iii. 54, where faithful 

VOL. II. PART III.—26 


380 A GRAMMAR OF NEW TESTAMENT GREEK. [8 157 


(applied to woman); or (6) from πιστός (« πίνω) -επτοτός, liquid. If a 
loan-word, John Lightfoot’s conjecture (Hor. Hebr. ii, 446), followed by 
Merx (ap. Mk 14%), is possible, that we have a transliteration of the 
Aramaic NPNDD, pistaca. So that the ointment was wnguentum balan- 
inum. Against this must be set the difficulty of the Syriac translator. 
Abbott (J.V. 252), following Wetstein, who quotes abundant instances 
of σπικάτον as the name of an ointment (<spica, cf. Vulg. spicati), 
suggests that an early Galilean tradition, finding in the original some form 
of σπικάτον, played upon it by saying “‘ not σπικάτον but πιστικόν.᾽᾽ 
W. C. Allen (Comm. on Mk. 168) supposes σπικάτον transliterated into 
Aramaic and misread by the Greek translator. (See also Vocab. s.v.) 


7. Nouns in -ίσκος, -ἰσκη. 


A frequent diminutive suffix, represented 


, , 
«ἰσκος, -ἰσκ : ; 
ΐ ae only by Τβασιλίσκος, νεανίσκος, παιδίσκη. 


Βασιλίσκος, which D reads in Jn 445. 4°, a diminutive precisely 
corresponding to our princelet. So Polyb. iii. 44. 5, Dittenb. OGIS 2011: 1° 
(quoted by Bauer τη loc.). 

Παιδίσκη. The deteriorative force of this diminutive is constant in 
the NT. Cf. Meyer, Ostr. 578 (A.D. 192) and Deissmann’s note, LAE 186 
(2200) n.7 “παιδίσκη meaning as in the NT a ‘ female slave.’ ” 

The formation of a diminutive νεανισκάριον (Epict. ii. 16. 29) shows 
that νεανίσκος was a “ faded diminutive.” 


(2) The -g- Suffixes. 
1. Nouns in -ay-, -vy-, -ἴγ-. 


This group was no larger in Greek than 
in the cognate languages; see Brugmann 
Grd.2 τι. i. 506 ff. “Apwag (for ἁρπαγή, see p. 335), πτέρυξ (for 
πτερύγιον, see p. 343), μάστιξ. 


-ay-, -υγ-, -ἴγ" 


2. Nouns wn -γγ-. 


A group of words denoting a hollow or 
oily a musical instrument is represented in the 
NT by λάρυγξ, σάλπιγξ, φάραγξ. 


(3) Lhe -χ- Suffixes. 


To this small miscellany belong στόμαχος, 


WE θρίξ, ὄρνιξ (see p. 130). 


§ 158-159] SUFFLXES. 5 381 
G, Stems in -c-. 
Stems un -os : -eo-. 


§ 158. (a) Nouns. For this class see Ὁ. 138 
above. Of the sixty-two nouns found in the 
NT notes are required for very few. 


τος (-εσ-). 


ΤΓλεῦκος first appears in Arist. For vernacular use of this NT ἀπ. 
Aey. in Ac 218, see Vocab. s.v. 

Ἔλεος and σκότος ; see above, pp. 126 f. 

Νῖκος. An old word, as Lobeck Phyrn. 647 shows. Wackernagel, 
Hellenistica 27, suggests that νεῖκος, a poetical word in Attic, and 
alive in Ionic for Herodotus (=contention), passed into the Kown with 
the meaning victory, through confusion with νίκη. 


Στρῆνος, This d.A. in NT at Rev 183 first appears in the New Com, 
See Lobeck Phryn. 381, Kennedy Sources 41. 


(b) Adyectives.—For this class see p. 162. 


THE FORMATION OF VERBS. 


§ 159. In considering the formation of verbs by suffixes we 
have to do with present stems, and shall follow the classification 
set forth above on pp. 184f. Under most of the headings a 
classified list of verbs occurring in the NT will suffice without 
further comment. Class VII., however, is very rich in types, 
and was specially productive in the Hellenistic period. These 
types must be discussed in greater detail. As a general rule 
the simplex preceded by a hyphen is given for composita. If 
the Preverbia do not follow within brackets, they may be found 
by reference to the List of Verbs, § 95. 


I. a. Person suffixes added to root. 
(a) With thematic vowel : 


-ἄγχω, ἄγω, ἀλείφω, ἀνοίγω, dpxw, βλέπω, βούλομαι, βρέχω, Bevo, 
γράφω, δέομαι (p. 195), δέρω, δέχομαι, διώκω, -δύω, -εἴκω, ἐλέγχω, 
ἕλκω, ἐμέω (p. 236), -ἕπομαι, ἐρείδω, ἐρεύγομαι, ἔρχομαι, εὔχομαι, 
ἔχω, ζέω (p. 195), ἥκω, θέλω, θλάω, θλίβω, θραύω, καθεύδω, λάμπω, 
λέγω, λείπω, λούω, λύω, μέλλω, -μέλομαι, μέμφομαι, μένω, -νέμω, 
νήφω, -οἴχομαι, παίω, παύω, πείθω, πέμπω, πλέκω, πλέω (p. 195), 
πνέω (ρ. 195), -πνίγω, mpiw, πταίω, ῥέω (p. 195), σέβομαι, -σείω, σήπω, 


382 A GRAMMAR OF NEW TESTAMENT GREEK. _ [ὃ 159 


σπεύδω, στήκω, στρέφω, THKW, τρέπω, τρέφω, τρέχω, τρΐβω, φείδομαι, 
φέρω, φεύγω, φθέγγομαι, -χέω (p. 195), ψεύδομαι, -ψύχω. 

Nore.—In συνθλάω the simplex θλα(σ)- ὦ may be from *dhysat>Skt. 
drsat, mill-stone (cf. δειράς), see Boisacq 347 η.". 

(b) Without thematic vowel : 

εἰμί, -εἶμι, ἐπίσταμαι, κεῖμαι, κρέμαμαι, φημί. 

Νοτε.---Κρύβω (περι is a late formation, following the analogy of 
rpi8w:érpuva. (See Thumb in Brugmann Gr.* 375 n. 1.) 


I, 8. Reduplicated forms. 


(a) Thematic : 


Γίνομαι (class. γίγνομαι, see p. 232), πίπτω (*7i-rre), τίκτω (*ri-TKo). 


(δ) Unthematic : 
Δίδωμι, -ἴημι, ἵστημι, κίχρημι, τίθημι ; (with nasal inserted) 
πίμπρημι. 


II. With formative suffix in -n-. 


a. (a) Suffix vo: ve or avo: ave. 


(i) Added to root: Adkvw, δύνω, ἱστάνω (p. 241), κάμνω, -κτέννω 
(p. 245), πίνω, τέμνω, -χύννω. Αἰσθάνομαι, ἁμαρτάνω, αὐξάνω, βλασ- 
τάνω (but see p. 231), Τὀπτάνομαι (for this late present, a back-formation 
from ὥφθην, see pap. instances in Vocab. s.v.). 

(ii) Added to root with nasal inserted: Θιγγάνω, λαγχάνω, λαμβάνω 
(p. 247), λανθάνω, λιμπάνω, μανθάνω, πυνθάνομαι, τυγχάνω. 

(b) Suffix νη : νᾶ added to root. Δύναμαι, ὀνίνημι (p. 251). 

B. (a) (i) Suffix vvo: vue. Δεικνύω. 

(ii) Suffix vFo: vFe. Τίνω, φθάνω. 

(iii) Suffix veFo : νεξε. -τἱκνέομαι, κινέω (<I.E. *gi-, οἵ. Lat. 
ac-citus), προσκυνέω (κυνέω -- ἔκυ-νεσὼω « ἕκυ-νε-σ-μι, according to Johans- 
son, followed by Boisacq. Brugmann (Grd.? 11. iii. 276), however, follows 
Wackernagel in deriving Skt. cwmba-ti ‘‘ kissed” from *cunva-ti, which 
supports F as against o). 

(b) Suffix νῦ: νῦ added to root. ᾿Αμφιέννυμι, δείκνυμι, ζώννυμι, 
κεράννυμι, -κτέννυμι (p. 245), -μίγνυμι, -ὄλλυμι (ξὀλ-νῦ-μι), ὄμνυμι 
(p. 251), πέταννυμι, πήγνυμι, ῥήγνυμι, ῥώννυμι, σβέννυμι, στρώννυμι. 


III. With formative suffix in so: se. 


Αὔξω (cf. Lat. aug-eo), κλά-(σ)-ω, σεί-(σ)τω, σπά-(σ)τω. 


§ 159-160} SUFFIXES. 383 


IV. Suffixes in sko: ske. 


(a) Added to simple stems: ᾿Αρέσκω, βόσκω, γηράσκω, μεθύσκομαι, 
πάσχω (--πάθ-σκω), φάσκω, -φαύσκω, -φώσκω (for relation between 
these last two words, and possible derivation of former, see above, p. 263). 

(Ὁ) Added to reduplicated stem: Βιβρώσκω, γι(γ)νώσκω, διδά(κ)σκω, 
-διδύσκω (ἐν-), μιμνήσκω, πιπράσκω. 

So also ἱλάσκομαι, according to Boisacq, p. 373, -«- ἔσι-σλᾶ-σκομαι. 

(c) With « before the suffix: ᾿Αναλίσκω (but see p. 228 above), 
γαμίσκομαι, εὑρίσκω, -θνήσκω. To these we may add the dz. ey. 
σταυρίσκω, Hv. Petr. ii. 3. 

Notr.—1l. According to J. Wright (Comp. Gram. 290), in such verbs 
as ἁλίσκομαι and εὑρίσκω the -.- was the weak grade form of an original 
long diphthong -é-, -d1. Such presents as Attic θνήσκω and μιμνήσκω 
were formed by analogy. 

2. The inceptive meaning which is so prominent in Latin verbs in 
-sco is rarely traceable in Greek. Μεθύσκω, which in the active is used as 
a causative of μεθύω, means, in the middle, “ to get drunk.’ But that 
cannot always be pressed, as 1 Th 57 shows. (See Milligan Commentary 
in loc.). Tauiox is used in a causative sense in Lk 203+, and is equivalent 
to γαμίζω in v.%8, 


V. Suffixes in to: fe. 


See above, p. 185, where it is shown that these may be ignored. 


VI. Suffixes in 00: θε. 


This small class is represented by Τζἀλήθω, ἔσθω, ἐκνήθω, νήθω (see 
LS), πλήθω, πρήθω. The present stems πλήθω, πρήθω are not 
found in NT. ᾿Αλήθω, κνήθω and νήθω are Hellenistic forms for ἀλέω, 
κνάω and véw (Rutherford NP 90, 134, 240). 


VII. Suffixes in yo: ye. 


§ 160. For this very large class of verbs it will be convenient 
to discuss word-formation under different headings from those 
of the divisions suggested on p. 185. 


i. Vocalie yo- presents. 


These consist of 1, dw; 2. éw; 3. ow; 4. iw; 5, va; 
6. evo, 


384 A GRAMMAR OF NEW TESTAMENT GREEK. _ [§ 160 


1. Verbs in ao. 

᾽ (a) A few root verbs, ἐάω (etym. uncertain, 
ae Boisacy, s.v.), ἰάομαι (-- Ἐἰσᾶ-ιο-), κτάομαι (cf. pf. 

κέ-κτη-μαι, κτῆ-μα, Skt. hsdyatt), and, with suffix -m-, κοιμάω (c. κοίτη, 

κεῖμαι). 

(Ὁ) Denominatives from ἄ- stems supply the majority. 

Bodw, δαπανάω, διψάω, θεάομαι, καταράομαι (<kardpa), καυχάομαι, 
κολλάω, ναρκάω (κατα-), νικάω, ὀδυνάω (see Vocab.), ὁρμάω, πεινάω, 
πειράω, σιγάω, σιωπάω, Τσπαταλάω (once in Polyb., in LXX and late 
writings), συλάω (<ovAn or σῦλον, both used mostly in plur.), τιμάω, 
τολμάω, τρυγάω, tpupdw, φυσάω (éu-), χολάω. To these may be added 
opdw for which Siitterlin (p. 10) postulates *épa(cf. φρουρά and Germ. 
*waro). 


(c) ᾿Αγαπάω, yevvdw, Τέἐραυνάω (see p. 86), ἡττάομαι, μεριμνάω, 
πλανάω are not derived from the corresponding nouns in -ἄ, which 
are back-formations or “‘noms postverbaux”’; see pp. 335, 356 supra. 
For ἐραυνάω see Solmsen Gr. Wortf. 50, who also (pp. 48 1.) derives 
μεριμνάω from ἕμερίμων <*pépimos (cf. μάχιμος etc.). Τεννάω, ἐρευνάω, 
and πλανάω had originally an -n- suffix, but were later taken over into 
the -6- conjugation. ‘Hrrdopa ( -- ἥττων) has replaced *jrrdopa (cf. Ion. 
ἑσσόομαι) under analogy of νικάομαι. (See further, p. 107.) 

(4) Conformity to type influences many verbs, especially when a 
considerable group already exists with the same general meaning. 

Thus the large group of verbs of sound in -dw is represented in NT 
~ by βριμάομαι (ἐμ-}} and μυκάομαι (<I.E. *mik-, extension of mu-, an 
onomatop. word, Siitterlin 25). Another drawn from agriculture, and 
represented in (b) above by τρυγάω, supplies us with ἀλοάω («ἅλως 
(Att.), Τἅλων), ἀμάω, λικμάω (-«λικμός). Sickness, and unhealthy desire, 
furnish another group, as χολάω in (Ὁ). This may possibly account for 
the LXX μοιχάω, NT μοιχάομαι (Xenophon and Hellenistic), which re- 
places the class. μοιχεύω (see Vocab.). 

Analogy accounts also for the following : ᾿Ατιμάω, Mk 124 Ὁ (<druos), 
follows τιμάω (<tiun): ἀντάω (ἀπ-, ὕπ-) <adv. ἄντα (see ὃ 114) conforms 
to the pattern πειράω: πεῖρα, so περάω (δια-) <adv. πέραν. Μωμάομαι 
(--μῶμος) may be influenced by λωβάομαι (λώβη). 

(e) A few verbs in -ἄω apparently come under none of these headings : 
᾿Αριστάω (<dpiorov), ἐμπιπλάω (pp. 205, 254), ἐμπιπράομαι, Ac 28° x* 
(p. 254), κολυμβάω (--κόλυμβος), μασάομαι (Aristoph. and Hellen.) 
<*paOsaopar <*pabsa, I.E. *mnth-ia (Boisacq, 8.v.), πηδάω (k=) «πηδόν 
<LE. *péd, which has the long grade of *ped, the root found in πέδη, 
πέζα, πέδον, πούς ; also χαλάω (deriv. uncertain, Boisacq, s.v.), ψηλαφάω 
(see Boisacq, s.v. ψάλλω). 

1 See § 163 (3) (c) below. The new LS maintains this distinction between 
βριμάομαι and “όομαι, as between sound and feeling. : 


§ 160-161] SUFFIXES. 385 


(f) Τ᾿ λεάω is a later form of ἐλεέω (p. 235, and for confusion of: 
flexions, pp. 195, 197, 198). Προσδοκάω, though simplex is δοκέω. 
TEMoydw, a Kow word, is an instance of a verb in -dw formed from a 
prep. phrase (see ὃ 118 and Vocab. s.v.). Γελάω (<*yedao-10, cf. γέλως) 
is an example of a consonantal yo- present. 


(g) There is really no justification for treating 
verbs in -rdw as a distinct class. The only possible 
example in NT of a frequentative force is in σκιρτάω (: σκαίρω), though 
it is doubtful whether there is any connexion between this ending and 
that of the Lat. frequentatives (e.g. dictare : dicere). As the nomina 
agentis in -rns regularly form their verbs in -τέω, that noun formation 
supplies no reason for a separate group here. The presence of τ in the 
stems of several d- nouns gives us Τβλαστάω (p. 231), μελετάω, 
teheuTdw. Φρεναπατάω (<dpevardrns) follows the analogy of ἀπατάω. 
Ἐρωτάω (<*épF-wr-) is connected with ἐρέω (épéFw) ; see Boisacq 278. 


-TAW 


(h) Some verbs in -caw are simply denominatives 
from stems in -a; eg. δειλιάω, κονιάω, Ἱπροαιτιά- 
ομαι (ἅ.λ. Rom 3%. Its simplex <airiais common in class. Gr.). Others 
are affected by the clearly marked groups in -τάω (Siitterlin 29 ff., and 
see (d) above), eg. Τἀγαλλιάω (for class. ἀγάλλω), ἀροτριάω (-- ἄροτρον), 
θυμιάω (to burn incense, distinguished from θυμόω, to be angry. The 
latter accords with the only meaning attaching to the Gr. word θῦμός, 
whereas θυμιάω goes back to the original and literal sense found in Skt. 
dhimah, 1.E. *dhii-mé-s), κοπιάω (-- κόπος), στρηνιάω (a word first found 
in the Mid. Comedy, see Lobeck Phryn. 381, Rutherford NP 475) 
<orpnvos. 


, 
“LAW 


(ἢ For fj, χρήομαι, the only two remaining verbs in the class -ἤω, 
see p. 195. 


2. Verbs in -éw. 


§161. The principal classification follows 
the distinction between verbs derived from 
simple noun stems, and those derived from compound nouns 
(and adjectives). In early Greek these verbs were more com- 
monly formed from simple nouns in -os. The proportion gradu- 
ally changed, until in the Hellenistic period the overwhelming 
majority of new formations came from compounds. 


-ἔω 


Siitterlin (p. 63) examines and corrects v.d. Pfordten’s tables, and 
with 1160 verbs in -έω shows the following ratio of new formations in 
(a) Homer, (6) Classical, (c) Post-classical authors— 

From simplicia, (a) 50, (ὁ) 30, (c) 10. 

From compounds, (a) 20, (b) 450, (c) 600. 


386 A GRAMMAR OF NEW TESTAMENT GREEK. _ [§ 161 


A. Verbs in -é from simplicia. 


(a) Denominatives in -é (I.E. *-e-76) corresponding to stems ending 
in -o-, where the -e- represents the °/, gradation. 

᾿Αθλέω (ἄθλος), ἀντλέω (fr. “Aos=hold of ship, then bilge-water) to 
bale out, thence simply to draw (water), dpyéw (Cyés<d-, ἔργον), ἀριθμέω 
(μος), αὐλέω (Ros), γαμέω (μος), δειπνέω (νον), Τδεσμέω (μος) read 
by CD and late uncials at Lk 8539 for δεσμεύω (NBL 33) late and rare, 
δωρέομαι (ρον) (LX X -éw, as less often in class.), θορυβέω (Bos), θρηνέω 
(‘vos), θροέω (Coos), καρτερέω (ρος “κάρτος, Ep. and Ion. for κράτος), 
κοινωνέω (“νός), κοσμέω (μος), λοιδορέω (pos), μετρέω (ρον), μιμέομαι 
(μῖμος), (παρα-)μυθέομαι (μῦθος), νοέω (νόος, νοῦς), νοσέω (σος), οἰκέω 
(*kos) (so ἐν-, κατ-, ἐνκατ-, συν- : but παρ- and περι- probably from 
cpds., see below, p. 389), ὀκνέω (°vos), ὁμιλέω (°Aos) (for meaning see 
Vocab. s.v., and for ὅμιλος, see above, p. 335), ὁμορέω (συνομορέω a ad. 
in Ac 187; συνόμορος is only found in later eccles. writers), from ὅμορος 
(see above, § 107, p. 284), is found as early as Hdt. and occurs in Plut. 
See also Syll.2 64118, ed.? 104416), ὀχλέω (“Aos), πατέω (- πάτος <*pnto-s 
[ef. πόντος, Skt. panthah, Lat. pons -tis] <I.E.*pent(h) : see Boisacq 803), 
πλουτέω (Tos), (ἐπι-)ποθέω (°Gos), πολεμέω (vos), (Sta-, KaTa-) -πονέω 
(vos), στοιχέω (°yos), (ἀπο-)στυγέω (γος), THPEw (possibly + from typos; 
which is only found in Aesch. Supp. 248. It may come from I.E, *q¥é- : 
ef. Skt. cdyati, “ perceive, watch” ; O. Slav. caja <*kéja, “ wait, hope” ; 
Boisacq, s.v.), ὑμνέω (‘vos), ὑστερέω (ρος), φθονέω (“vos), φιλέω (dos), 
φλυαρέω (“pos), φοβέω (°Bos), φρουρέω (°pos), χωρέω (“pos), ὠνέομαι (vos). 

Notre.—l. αἰνέω (aivo-s) belongs to this group, as Hom. ἤνησα shows. 
The later form ἤνεσα (cf. fut. αἰνέσω) is due to its antonym νεικέω 
(<vetkos, neut.). 

2. Novéw is denominative from ἔποι- ό-ς (so Boisacq, s.v.), which, as 
Siitterlin (p. 41) observes, only survives in cpds., 6.9. apromotds, baker. 


(b) Afew deverbativa in -éw (I.E. -é-16) with traces of frequentative 
force. Τγρηγορέω, a Hellen. back-formation from pf. ἐγρήγορα of ἐγείρω, 
πορθέω (<répOw), ῥιπτέω (<pirra, cf. iacto: iacio: see p. 257, also 
Brugmann KVG 536), σκοπέω (acc. to Brugmann-Thumb Gr. 360, this 
corresponds to σκέπτομαι as an iterative. See above, p. 258, for the 
denominative ἐπισκοπέω in 1 Pet 5® (Aw)), φορέω, frequentative of φέρω 
(but see Debrunner Worth. 95 for popéw< opos). 


(c) Denominatives in -έω (<*-eo-s0) from s- stems. A few of these 
preserve traces of their origin outside the present tense, e.g. τελέω, aor. 
ἐτέλεσα, f. τελέσω. Most of them have conformed to the *-e-76 type. 


1 Mr. E. E. Genner observes that the word τηρός in Aesch. Supp. 248 is 
almost certainly corrupt. [Sidgwick’s textual note in the OCT is “ἢ τηρόν 
ex ἤτηρον ut videtur factum M.”—Ep. ] 


§ 101] SUFFLXES. 387 


( Απε)ἀλγέω, (ἐπι-, κατα-)βαρέω, a later form of βαρύνω, ἐλεέω (see 
above, pp. 195-7, 235, 385), θαμβέω, θαρρέω later form of θαρσέω, κρατέω, 
μισέω, πενθέω. All these are formed from neuter nouns in -os, -eo-. 

Note.—Bapéw and θαμβέω are Ionic contributions to the Κοινή.} 


(4) Denominatives from other stems. 

᾿Απειλέω (ἀπειλή), [ἐλαττονέω (ἐλάττων). See Vocab. s.v. for other 
occurrences of this rare word. ‘lotopéw (ictp<*Fid-rop-). For the 
history of the meaning of this word (d.A. in NT Gal 115 =o visit) see Burton 
ICC in loc., 59 Vocab. sv. Λυπέω (: λύπη following ἀλγέω : ἄλγος), 
μαρτυρέω (μάρτυς, -υρ-), φρονέω (φρήν, φρεν-), Pwvew(: φωνή, following 
ἀὑὐτέω : av’rn, and other primary verbs of sound). 

(6) A number of verbs in -έω remain, which are not formed from 
compounds, but for which a Greek simple noun is not quotable. Some 
are root verbs, but not all: 

Aipéw, ‘“‘létymologie de αἱρέω est incertaine; Vinitiale n’a pas été 
F.” Boisacq, s.v. 

Αἰτέω, formerly connected with αἰκίζω, Skt. ydcati, Goth. aihtrén, and 
derived from base ateg’. For Buck’s phonetic objections to this, 
see Boisacq, 8.0. 

Τ᾽ Αλισγέω, a late verb (LXX*”), is attested by Τἀλίσγημα (Ac 152°), 
Boisacq notes “le groupe -σγ- est énigmatique ; rapport possible 
avec ἀλίνειν.᾽ 

᾿Αρκέω, cf. Lat. arceo, αἴ, arcdnus, from base areg “ to avert,” ‘“ repel.” 
Brugmann (Ογά.3 11. iii. 339) derives from Skt. rakga-ti “to arm,” 
‘deliver’? <*(a)leq-s(o)-. 

"Apveopat, possibly connected with Arm. uranam “to refuse,” from *6r. 

*Ackéw. Uncertain etym. Boisacq quotes theory deriving word 
from *d-oxds<*av-oxos, and eft. ἀνακῶς carefully « κοέω. 

Τ᾿ Ἐνειλέω (Mk 154%, εἰλέω Hv. Petr. vi. 24) for classical ἐνείλλω (Thuc.). 
See Vocab. s.v. 

Ζητέω « ἔδιάτειω. Cf. Skt. ydtati “to fix,” ydtaté ‘to make an effort.” 

Καλέω, weak and strong grades seen in καλέ-σαι, κέ-κλη-μαι, Lat. 
calé, -dre, clamor, clamo. I.E. *qald- and *gel (-. See Boisacq, s.v. 


1 See Thumb DAC i. 555a. ‘‘ Words like ἀπαρτίξζω (in ἀπαρτισμόΞ), ἔκτρωμα, 
κοπάζω (of the wind), ὄλυνθος, σανδάλιον, σκορπίζω, ete., in the LXX or NT 
are of Ionic origin. The Ionic element includes, further, the so-called poetical 
words of the Kow7, i.e. Hellenistic words which formerly were to be found only 
in the poets, but which from the fact of their occurrence in papyrus texts 
concerned with matters of everyday life, and partly also from the fact of their 
survival in MGr, are now seen to have belonged to the colloquial language. 
They include, ¢.g., Bapéw, ἐντρέπομαι, θαμβέω, μεσονύκτιον, πειράζξω, ῥάκος, ὠρύομαι, 
in the LXX and the NT, and ἀλέκτωρ, βαστάζω, ἔριφος, φαντάζω, φημίζω, in the 
NT. Words of this class were imported first from the literary Ionic of the 
earlier period into the language of poetry, and then again from the vernacular 
Tonic of the later period into the Kow7, and there was no direct link of con- 
nexion between the two processes,” 


388 A GRAMMAR OF NEW TESTAMENT GREEK. [§ 161-162 


(Ex=)xevréw to prick, stab, goad, of. xovrés pole (Lat. contus), Skt. 
cnathati, cnathayati, “to pierce.” 

Λᾶκέω (see p. 246 and Vocab. s.v.), ef. Lat. loguor (<*laquor, *laq-uo-). 
I.E, */’g-, extension of */’, reduced form of */d(z)- “‘ to cry.” See, 
further, Boisacq, s.v. λαίειν. 

Aahéw, onomatop. cf. Skt. lalalla. The original sense of to chatter 
had quite left the word in NT times. See Vocab. s.v., where a 
number of exx. from papp. “ all bear out the usual distinction that 
while λέγω calls attention to the substance of what is said, the 
onomatopoetic λαλέω points rather to the outward utterance.” 

Muéw from ,/ pi, wv, a sound made with closed lips. J. A. Robinson, 
Ephesians 234, casts doubt on the derivation from μύω as that, 
when used simply, always means to close the eyes, not the lips. 
For the fading of technical meaning from μυέω and μυστήριον in 
later Greek, see Vocab. s.v. 

"Opxéopar. Cf. Skt. rghdydti ‘to tremble,” I.E. *ergh- (see Boisacq, s.v.). 

Πτοέω from ./ πτω-, rra-, πετ- <‘*pet- seen in πέτομαι, πίπτω, in πτάξ, 
πτώξ, πτώσσω (See Boisacq 823,—also Siitterlin 84, on late date 
of πτοία from which verb has sometimes been derived). 

Πωλέω «ΤῈ. *pel-. Cf. Skt. panaté (*prnaté <p]-n-). 

(‘Atro) crepéw. Root uncertain. Boisacq cft. MIv. serbh “‘ theft.” 

(Απ-, ἐξ-) ὠθέω. Cf. Skt. vadh- “to strike,’ <IE *uedh- *uodh- *uddh- 
(see Boisacq, s.v.). 

᾿Ωφελέω (see above, ὃ 111, and Boisacq 732, 1085). 


B. Verbs in -έω from compounds. 


§ 162. (a) Denominatives corresponding to stems in -o-. 

Many of these are verbs in -φορέω, -ποιέω, -oyéw, -εργέω, which were 
very productive verbal endings in later Gr. (For tables of relative 
frequency in class. and post-class. writers, see Siitterlin 49). 

The most noticeable groups in the NT vocabulary are those in— 


-αγωγέω : δουλαγωγέω, Τσυλαγωγέω, ᾿χαλιναγωγέω (only Jas”®’, Lucian®*: 
γος first appears in Chryst.), Τχειραγωγέω (Ps-Anacr., LXX, 
Ev. Petr. x. 40). 

-αρχέω : πειθαρχέω (ὃ 108, p. 290) represents for the NT this very large 
group, whilst for ἡτετρααρχέω, which has χης for its cognate in 
NT, we may either postulate a form in -os, or account for the 
verb by analogy (see Siitterlin 79). 

-yovéw: ζωογονέω (see Vocab. s.v. for LXX and NT meaning, to pre- 
serve alive, as against class. use=to endue with life, Ἷ τεκνογονέω. 

-δημέω: ἀποδημέω (ὃ 115, p. 299), ἐκδημέω, ἐνδημέω, ἐπιδημέω (§ 120, 
p. 315). 

-ἐργέω : Tayaboepyéw (1 Ti 618: the rare contracted form dyaboupyéw 
appears in Ac 147), ἐνεργέω (ὃ 118, p. 308), συνεργέω, γεωργέω 


§ 162] SUFFIXES. 389 


(§ 105, p. 271), Ttepoupyéw, λειτουργέω (§ 106), συνυπουργέῳ 
(elsewhere found only in Hipp. Art. 824, and Lucian Bis Accusat. 
17) is an instance of the tendency in later Gr. to coin double epds. 
Ὑπουργέω ( --  γύς) is common in class. Gr., though “yos is not found 
before Xenophon. 

-ηγέω : ὁδηγέω, χορηγέω (see ὃ 105, p. 275). 

-ηγορέω : ἰἀλληγορέω (<°pos, ace. to ν. d. Pfordten Gr. Denom. 35, but, 
ὅρος not given in LS.!_ For verb see Burton Gal (ICC) 254 f.), 
δημηγορέω (§ 105, p. 273), κατηγορέω. 

-θετέω : ἀθετέω (see Vocab. s.v.) is from τος, with which we may com- 
pare νομοθετέω (<°rns) and νουθετέω, for which a primitive ‘ros 
is not extant. 

-θυμέω : ἀθυμέω, εὐθυμέω (ὃ 118, p. 308), Ἱμακροθυμέω. 

-λογέω : ἀπολογέομαι may possibly come in here, but see § 115, p. 299, 
and Siitterlin 52. j}Batradoyéw (ὃ 105, p. 272), γενεαλογέω 
(found as early as Hdt., whereas “yos is not quotable until Dion. 
H. Cpds. formed with λογέω were very numerous, so that analogy 
probably plays its part), edAoyéw (given by v. d. Pfordten 54 as 
from °Aos, but as the meaning of the adj. is reasonable, and of the 
verb to praise, bless, the cpd. of ev and λέγειν may simply follow 
the common type of -λογέω verbs, esp. its antonym), κακολογέω 
(<°yos), ὁμολογέω (ὃ 107, p. 284), dppodoyéw (ἴσυν- see § 105, 
Ρ. 272), Ἰστρατολογέω (2 Ti 2*). 

-μαχέω : in ᾿θυμομαχέω and λογομαχέω (ἅ.λ. 2 Ti 214) the first constituent 
is instrumental (see ὃ 105, p. 273), whereas in θεομαχέω (Ac 23° 
HLP) and Τθηριομαχέω (<°os not “ns, see Siitterlin 79) it is 
dative. 

-νομέω : KANPOVOLEW, οἰκονομέω, παρανομέω. 

-οἰκέω: Unlike the οράβ. of οἰκέω mentioned above, παροικέω and 
περιοικέω are derived directly from πάροικος, περίοικος. This 
is clearly seen in the former by the changed meaning of the verb 
in Hellenistic corresponding to the changed meaning of the adj. 
from neighbouring to foreign, alien. (For πάροικος in Hellenistic 
see Deissmann BS 227 f.) 

-roéo: ἀγαθοποιέω, Τεἰρηνοποιέω, ζωοποιέω, κακοποιέω, TKadoTroréw, 
Ἱμοσχοποιέω (ἁ.λ. Ac 75), Τὀχλοποιέω (ἅ.λ. Ac 175). A special 
note may be given to ὁδοποιέω (from Xenophon), which is read 
in Mk 2% by BGH 13 etc. It is tempting to adopt this rdg. with 
WH mg., and so avoid the incorrect use of the active for the 
middle. But the avoidance of the term in the Synoptic parallels, 


1The new LS cites the Htymologicum Gudianum, 515. 42. Mr. E. E. 
Genner tells me that this medieval glossary, based on material of the early 


Byzantine age, quotes, s.v. συνήγορος, παρήγορος and ἀλλήγορος as parallel 
forms, 


390 A GRAMMAR OF NEW TESTAMENT GREEK. _ [§ 162 


and the evidence of confusion between act. and mid. in the papyri 
(see Prol. 159), raises a doubt. For the word see Vocab. s.v. 

-πορέω : ἀπορέω, εὐπορέω, ὁδοιπορέω. 

-τομέω : διχοτομέω (ὃ 106, p. 281), Ἰλᾶτομέω (elsewhere in LXX, Justin 
M., Diod., «μος [- λᾶς, τέμνω] only found in ΤΙΧΧ and Josephus), 
Τὀρθοτομέω, a direct formation on analogy of other epds. in -ropéw. 
See above, § 105, p. 274. 

-φημέω : βλασφημέω (ὃ 105, p. 272), δυσφημέω (§ 107, p. 287). 

-popéw: εὐφορέω (<"pos, which started with the passive sense of bear- 
able, and then developed the active meaning, fruitful, produc- 
tive, common in the medical writers (Hobart 144), and in 
Hellenistic. See further Siitterlin 42), καρποφορέω, fm\npo- 
φορέω, which follows the analogy of τελεσφορέω (ὃ 105, p. 275), 
tpotropopéw (Ac 1318 SBC*Dw) also found as a ν.1. in Deut 13: B*; 
its existence (=épw τὸν τρόπον) is attested by Cicero Aft. xiii. 
29.2. In the absence of “pos we must regard the verb as a direct 
formation. +Tpododopéw (tb. AC*E 33) is the reading of B*AF in 
Deut 151, and the word occurs without v./. in 2 Mac 757, The adj. 
“pos is not found before Eustathius. 


A number of these verbs in -ἔω were formed from noun compounds 
having a- privative or εὐ- as a prefix. In addition to those given above, 
we find in the NT dyvoéw (<*a:yvoo-s, <*yva-Fo-s, cf. Lat. cognitus, 
<*-gna-to-s, see Brugmann Grd.” i. 203; but see above, ὃ 106, p. 281), 
ἀδικέω (as in class. Gr. both intrans. and trans., whereas Τἐκδικέω (<°kos) 
is only used transitively. The latter verb is only found in Kowy. For 
papyrus exx.see Vocab.). ᾿Αδυνατέω (Xen., Plat., Arist.) from τος has given 
us the analogous fduvatéw from °rés, a verb found in Philodemus the 
Epicurean philosopher of i/B.c., elsewhere only in Paul (Rom 144, 2 Co 9° 
138). Τ᾽ Ακαιρέω (Diod. Τπτέομαι, N.T. aA. Phil 44%, Herm. Sim. ix. 10°) 
is a Hellenistic derivative from the class. ἄκαιρος, opposed to Τεὐκαιρέω, 
which is a good Κοινή word (Polyb., Plut., papp.) condemned by 
Phrynichus and Photius (Rutherford NP 205), who prefer εὖ σχολῆς ἔχειν 
(ὅρος and ‘pita are sound Attic, but not in the sense of σχολαῖος and 
σχολή). See Vocab. s.vv., as also for remaining words in this group, 
amotéw, Τἀστατέω and ἀτακτέω. ΤΕὐαρεστέω is used by Hellenistic 
writers alone (so “tos can be quoted from inscrr. and papp. in addition 
to the “ bibl. and eccl.” citations. Vocab. s.v.). Εὐνοέω (ὃ 107, p. 287). 

The remaining verbs in this class are : 

᾿Αγραυλέω (§ 107, p. 283), ἀγρυπνέω (δ 108, p. 290), αἱμορροέω (<°poos, 
where the first element in the word is instrumental in case relationship ; 
see T.P. cpds., ὃ 105), ἀκολουθέω (§ 107, p. 285), Τἀντοφθαλμέω (unless this 
vb. should be placed in class (e) below. See above, ὃ 114 (6), p. 297), 
βραδυπλοέω (<**r)oos, see ὃ 107, p. 284), διακονέω (<*vos, but see ὃ 116, 
p. 303, for another possibility), ἐπιορκέω (ὃ 120, p. 314), ἑτεροδιδασκαλέω 
(§ 107, p. 284), Τἑτεροζυγέω (<*yos) occurs first in 2 Co θ᾽: (see § 107, 


§ 162] SUFFLXES. 391 


Ρ. 284), Τεὐθυδρομέω (Philo and NT; for “pos see ὃ 106, p. 282), Τεύπρο- 
σωπέω (ὃ 107, p. 287), εὐχαριστέω (for meaning of this verb in Hellenistic, 
see Milligan, Thess. p. 5, and Vocab. s.v., also Deissmann LAH? 
132 n.8, 168 n.? (7135 π,8, 179 n.°)), Τεὐψυχέω (§ 107, p. 287), εὐωχέω 
(συνευωχοῦμαι, first in Arist. See ὃ 106, p. 282), (ἀνατ)ζωπυρέω 
(<¢wrvpor, ὃ 107, p. 284), θεωρέω (<“pos <*Oea-opos, *-Fopos, cf. ὁράω, 
O.E. warén, ware), ἱεροσυλέω, Τκατακληροδοτέω (Ac 1319 minusc. pauc. 
for “νομέω ‘from missing active sense of kAnpovop.,’ (Knowling ΠΟΤ 
in loc. See Vocab. s.v. for use elsewhere), Τκληρονομέω, Τλιθοβολέω, 
μεγαλαυχέω (Jas 3° SC?KL, written divisim in other MSS ; both verb and 
°yos as early as Aeschyl.), vavayéw (for ναυᾶγός see ὃ 105, p. 274), ξενοδοχέω 
(spelt in Attic with « for x, as all the cognate words, see Rutherford 
NP 362), οἰκοδομέω, ὀλιγωρέω (for “pos see ὃ 107, p. 284), στενοχωρέω, 
ταλαιπωρέω (ὃ 106, p. 282), τιμωρέω, φιλοτιμέομαι (for these φιλο- and 
τιμο- cpds., see § 108), χειροτονέω (ὃ 105). 

(b) Denominatives corresponding to nomina agentis in -ns, -της. 

There is one doubtful example only of verbs formed from first de- 
clension nouns in -ns, τετρααρχέω (see under (a) above). 

As already seen in § 150 above, noun compounds show a strong pre- 
ference for the ending -rys when forming nomina agentis. Eleven of 
these supply verbs in -έω in the NT. In most cases the nouns have been 
discussed above, and the references are accordingly given: ᾿Αγανακτέω 
(§ 106), Ταὐθεντέω (ὃ 106), Τγονυπετέω (δ 105), εὐεργετέω (ὃ 106), νομοθετέω 
(see (a) above), Τοἰκοδεσποτέω (ὃ 105), πλεονεκτέω (ὃ 105), {προσω- 
πολημπτέω (ὃ 105), συκοφαντέω (§ 105), ὑδροποτέω, ὑπηρετέω ( 129 (b)). 

(c) Denominatives from compound adjectives in -ης (-es-). 

But for the absence of any survival in the s- conjugation, we might 
relate this group to the *-es-go- class in A (c). As it is, we can infer 
that before the Homeric period these verbs conformed in all ways to the 
predominating class of -o- stems. 

᾿Αμελέω, ἀπειθέω (for meaning, see Vocab. s.v.), ἀσεβέω and its opposite 
εὐσεβέω, ἀσθενέω, εὐλαβέομαι, κακοπαθέω (first in Xenophon), and the 
other -παθέω cpds., Ἱμετριοπαθέω (§ 107) and συνπαθέω (ὃ 127), λυσιτελέω 
(§ 108), διασαφέω, which appears to have been formed straight from σαφής, 
without the intervention of ἔδιασαφής (δ 116). 


(d) Denominatives from adjectival compounds in consonantal stems. 

Of stems in -ov- the -fpovéw cpds. are as old as Homer. Jn NT this 
formation supplies παραφρονέω (ὃ 123 (b)), σωφρονέω (§ 107, pp. 284-5), 
ὑψηλοφρονέω (ὃ 107). In addition there are two d- cpds., ἀδημονέω 
(§ 106), doxnpovew. 

Of stems in -p- ψευδομαρτυρέω ( 106). 

Of stems in -ὃ- Τὀρθοποδέω (ὃ 107). 

(e) Compound verbs in -έω formed directly on the model of those 
grouped under (5). 


392 A GRAMMAR OF NEW TESTAMENT GREEK. _ [§ 162 


This flexibility in verb formation goes back to very early times, for 
Cwypéw and ἐπιχειρέω appear in Homer and ἐνθυμέομαι is very common 
in class. writers. Still greater freedom was shown at a later stage in the 
language, for the remaining seven verbs in this list are purely Hellenistic. 


Τ᾽ Αντοφθαλμέω (Polyb., see Capes, Achaean League, p. 262), possibly a 
Ciass VII. verb? from ἀντ᾽ ὀφθαλμῶν (ὃ 114 (5)). 

᾿Ενθυμέομαι, from ἐν θυμῷ (ἔχω), see ὃ 118 (c), and, for meaning, Vocab, s.v. 

Τ᾿ Ἐνκακέω (Polyb.) for ἐν κακῷ εἰμί. See ὃ 118 (c) for derivation and 
meaning. 

Τ᾽ Ἐξουδενέω (LX X and pap.). See above, §§ 46, 119, and Vocab, s.v. 

᾿Επιχειρέω, a verbal epd. of Class VII. above (ὃ 109) -- χεῖρα θεῖναι ἐπί τι, 
without the intervention of a noun cpd.: see § 120 (c). 

Τεὐδοκέω (Polyb., Diod., LX X), a new verb made with an adv. (p. 292). 

Zwypéw, Class VI. above, from ζωὸν ἀγρεῖν (δ 108). 

ἹΚακουχέω (Plut.t LX X’s, common in papp., esp. in marriage contracts, 
see Vocab. s.v., though the derivative noun “va is found as early as 
Aeschyl.) seems to belong to Class VI. A (§ 108). 

tT Texvotpodpew (1 Ti 5'°. Elsewhere Arist., Epict.), Class VI. A. 

ἸΧρονοτριβέω (Ac 20'%, Elsewhere Arist., Plut., and late writers), Class 
VI. A. 

Notr.—The distinction of class in these verbs in -ἔω corresponds 
broadly to a difference in meaning, which again lies in the nature of the 
nominal types from which the verbs are formed. The compounds are 
adjectival in meaning, and the derivative verbs have the force of “ to be 
what the adjectival word stands for.’ The simple nouns, on the other 
hand, stand rather for things or abstractions. Other compounds are 
virtually nomina agentis, and the derived verbs have what at first sight 
appears to be a factitive force, though it actually comes under the general 
meaning “το be so and 50. The distinction is not absolute, and the 
following considerations may be observed. 

1. Verbs in -ἔω from simplicia sometimes admit the meaning “‘ to be 
what the noun represents.” £.g. in class. authors, διακονέω, καρτερέω, 
κοινωνέω, λαλέω, λοιδορέω, μιμέομαι, σκοπέω, τηρέω, ὑστερέω, φλυαρέω. 
Siitterlin 49 attributes this in τηρέω and σκοπέω to their being primitive 
formations, whereas φλυαρέω is from a nomen actionis “pos. In NT we 
have also δυνατέω and its predominantly Hellenistic negative ἀδυνατέω. 

2. The factitive appearance of some verbs in -éo from cpds. involves 
a question of accent. A good example is Τλιθοβολέω. The rule is that 
when a T.P. cpd. is a transitive, or active, verbal in -os, it accents the 
penult if this is short, otherwise the last syllable. But if the last part is 
intransitive, or passive (in meaning), the accent is recessive (see Goodwin, 
Gr. Gr? 194). Thus λιθο-βόλος thrower of stones, λιθό-βολος pelted 
with stones. The verb corresponds to the former only. Similarly 
διχοτομέω follows the active meaning of διχοτόμος. Of course this 


1 The classification here referred to is set forth in ὃ 102 above. 


5 


§ 162-163] SUFFIXES. 393 


uncertainty as to the accent of the nominal cpd. sometimes leads to 
ambiguity in the force of the derivative verb. 

3. A number of Possessive (B.V.) cpds. give rise to ambiguity in their 
derivatives. Thus, εὐθυμέω in class. writers is both trans. and intrans. ; 
in NT always intrans. Εὐπορέω class. both trans. and intrans. ; in NT 
intrans.=to be well off. Similarly ταλαιπωρέω in its N'T occurrence 
(Jas 4°) has the intrans. meaning, though exx. can be quoted from class. 
writers and LX.X to support an active sense =to weary, distress, 

4. Occasionally an unambiguous adjective results in a verb with 
double meaning. Thus στενόχωρος narrow gives “péw=(1) intr. to be 
straitened (LXX), anxious (Hipp.); (2) trans. to straiten, compress 
(LXX, Diod., papp.). So in NT the pass.=to be straitened. (See 
Vocab. s.v.) 

Similarly with simplicia. OdpuBos, wproar, gives “Béw; (1) to make an 
uproar ; (2) to throw into confusion. 

5. In Hellenistic there are numerous examples of a tendency, already 
traceable in the class. age, to give an active sense to the verb, though 
its corresponding adjective is passive. Thus the good Κοινή verb Τἀθετέω 
to annul, cancel, comes from ἄθετος, which is found in the passive sense 
null, void, set aside. 

6. Note the transitive force produced by the perfectivising preposition 
ἀπο-, in Τἀφυστερέω, to keep back (Jas 54), see ὃ 114. 

7. The deponent verb φιλοτιμέομαι may owe its form to the in- 
herently middle force of the word in its classical (but not Hellenistic) 
sense of to be ambitious ; or, as Siitterlin (p. 44) and Debrunner ( Wortb. 99) 
suggest, it may follow the example of other verbs of desire, e.g. βούλομαι, 


ὀρέγομαι. 
3. Verbs in -ὀὼ. 


P § 163. This class of verbs, if not peculiar 
ow . 
to Greek, must have arisen at a very late stage 
in the I.E. parent language. We have already seen that a- 
stems gave rise to -aw verbs, and -o- stems to -éw verbs. 


The origin of the -όω verbs may probably be traced to three or four 
influences. 


(a) The analogy of the instrumental -aw verbs would play a part. 
Thus, πέδη a fetter, πεδάω to fetter, σκέπη a cover, σκεπάω to cover, from 
which the transition was easy to στέφανος, στεφανόω. This tendency 
would be strengthened by the parallel formations μῆνις : μηνίω, μέθυ : 
μεθύω. 


(0) Apart altogether from the denominative verbs and their tense 
system, there was a direct formation of denominative adjectives, e.g. Lat. 
barba : barbatus, aeger : aegrétus. So in Greek we have κοντός : κοντωτός 
provided with a rowing pole (though this adj. is not attested before Diod.). 
Brugmann (KVG@ 532) postulates μισθός. : μισθωτός, and derives from this 


394 A GRAMMAR OF NEW TESTAMENT GREEK.  [§ 163 


μισθωθῆναι; μισθώσω. ἐμίσθωσα, and finally μισθόω. (See also Grd.2 τι. iii. 
206 and Gir. 357.) 

(c) Where there are pairs of nouns derived from the same root and 
closely related in meaning, one ending in an -d- stem and the other in an 
-o- stem, a verb in -ow may well have arisen from the noun in -os corre- 
sponding to the already existing verb in -aw. Thus χολή : χολάω, 
χολός : χολόω. (See Siitterlin 99.) 

(d) Dr. Giles finds the beginning of this series in -ow ‘‘ with denomina- 
tives like ῥιγόω from *pryas (gen. Ἔῤιγόος, cf. Lat. rigor), ἱδρόω from 
ἱδρώς (=*suidros, cf. Lat. stidor=*suoidds)’ (Manual* 442 n.). 


Whatever the origin, this type became very common in 
forming verbs from -o- stems with a factitive or an instrumental 
meaning. There are 96 verbs in -dw in the NT, together with 
25 additional compound verbs. Of these about three-fourths 
are derived from o- stems (in about equal proportions from 
nouns and adjectives), 7 from a- stems, 9 from 3rd Decl. 
nouns, 4 from adjectives with consonantal stems, and a few 
from prepositional and other phrases, 


(1) Denominatives from -o- stems. 


Many of these were normal in the classical period, some few 
are only found in the late classical writers, others first appear 
in prose in the Κοινή, whilst the considerable batch of fresh 
formations shows that this suffix was actively creative. Atten- 
tion need only be called to the following :— 


Τ᾽ Ακυρόω (<éxupos common in legal phraseology. Vocab. s.».). 

Τ᾿ Αναλόω. (Back-formation, see p. 228.) 

Τ᾽ Ανακαινόω (first in Paul, for class. “(fw (as in Heb 66 and LXX). See 
Vocab. s.v.). 

Τ᾽ Αναστατόω. (See§113. A vernac. word found in LX X, NT and papp. 
Vocab. s.v.). 

Τ᾽ αφυπνόω. (Late verb. In Anth. trans. to wake from sleep. Elsewhere, 
as in Lk 8%, to fall asleep. For this meaning see ὃ 115 and 
Vocab. s.v.) 

TBeBynd\ow. (Firstin LXX. See Vocab. s.v. “dos.) 

Ἰδεκατόω. (Siitterlin 108 would derive this from ἡ δεκάτη (μερίς), but 
the adj. in -ros would account for the Hellenistic suffix -d in 
place of class. °revw. Note, however, ‘‘ the rare ἀποδεκατόω (with- 
out var. Mt!, Lk!, Heb!) is replaced by the rarer ἀποδεκατεύω 
(N*B), Lk 1813 (WH App.” 178). A reason for the new coinage 
is suggested in Vocab. s.v. : 


δ 163] SUFFIXES. 395 


Τ᾿ Ἐντυπόω. (Hellenistic, though the simplex appears in Plato.) 

Τ᾿ Ἐπιδιορθόω. (Almost peculiar to Tit 15, is the common διορθόω epded. 
with émi-=in addition. See ὃ 120.) 

Θεμελιόω. (First in Xen. In MGr θεμελιώνω.) 

Θυμόω. (Act.in LXX only. In class. and in the one NT occurrence, pass. 
=to be angry.) 

Tilkavéw. (Act.in NT. Elsewhere pass., e.g. P Tebt i. 20°.) 

ΤΚατιόω. (Apart from Jas 5° only found in Sir. and Epict.) 

+Kauadw (for class. καυματίζω. See Mayor, ap. 2 Pet 339) 

Κεφαλαιόω. If this rdg. is right in Mk 124 (ἐκεφαλαίωσαν), κεφάλαιον has 
given rise to a verb with a totally different meaning. Lobeck 
(Phryn. 95) points out that κεφαλή 1» κεφαλίζειν, caput percutere, 
λαιμός >Aauiferv, ῥάχις  ῥαχίζειν and adds “alia huius 
significationis terminatio est in verbis γναθοῦν, 1.6. εἰς γνάθους 
τύπτειν, γυιοῦν, κεφαλαιοῦν in Hv. 1). Marci xii. 4.” But as 
κεφάλαιον does not mean head, but sum total, or chief point 
(whence ἀνακεφαλαίοω, fo sum up, see J. A. Robinson Ephes. 
145), there is much to be said for the rdg. of NBL, ἐκεφαλίωσαν. 
The verb would then be Tkepadidw, formed from κεφάλιον, a 
Hellenistic diminutive of κεφαλή. For further suggestions se 
Vocab. 8.v. 

Κημόω. Only in Xen. before Paul, who in 1 Co 9° (κημώσεις B*D*FG, 
φιμώσεις NACw) substitutes this verb for φιμόω, which our MSS 
of the LXX read in the pass. cited. (Lietzmann (HNT in loc.) 
thinks Paul here gives the true text of the LXX.) 

TKodoBow (first in Arist.), from xcodoBds=maimed, mutilated. For ver- 
nacular use of cognates of this verb see Vocab. s.v. 

Λυτρόω. Class. but well established in vernac. See Vocab. s.v. 

ἹΜαταιόω. LXX and Paul (Rom 15). Act. only in Jer 231°, where 
intrans. (=to pretend). 

tNexpdw. See Vocab. s.v. νεκρός. 

Παλαιόω. In act. peculiar to LXX and NT, but pass. in Hipp., Plat., 
Arist. 

Πωρόω. Factitive verb in medical and Hellen. writers. Metaph. meaning 
peculiar to LXX and NT. (For confusion with πηρόω in some 
MSS, see J. A. Robinson H'phes. 271.) 

‘Puréw. A rare class. verb given in NT lexicons on strength of T.R. in 
Rev 224%. Ῥυπωσάτω seems to be without MS warrant. 
According to vy. Soden and R. H. Charles the alternatives 
are puravOnrw (NX, 94, 2017), ῥυπαρωθήτω (205), ῥυπαρευθήτω 
(046 εἰ al.) ῥυπασάτω (2029). 

Σαρόω. For caipw, which was in common use in Tragedy. But even 
this was condemned as un-Attic by Phrynichus (Rutherford 
NP 156), who requires παρακορέω, a8 κόρημα for cdpov. Lobeck 
(Phryn. 83) remarks “ σαροῦν improbat Phryn., non caipeww.” 
VOL. II. PART III.—27 


396 A GRAMMAR OF NEW TESTAMENT GREEK. _[§ 163 


The explanation may be that capow had completely displaced 
σαίρειν =sweep by this time, the latter being only still used in 
the sense of to grin. 

ἸΣημειόω. For σημαίνω (2 Th 312, See Milligan in loc.). 

Σπαργανόω. From σπάργανον a swathing-band. Lk is preceded by the 
medical writer Hipp. as well as by Arist. It is also used by Plut. 

TEmAdw. A Hellenistic derivative from the word σπίλος, which in 
late Gk. came to mean a stain (Rutherford NP 87 and Lobeck 
Phryn. 28). 

Σταυρόω. In class. to fence with a palisade; in Polyb. and NT =to crucify. 

Etepedw. A factitive verb, not found before Xen. 

ΤΤαρταρόω. d.d.in2 Pet 24, though the cpd. xara’ is found in Sext. Emp. 

and other late writers. 

TdpayedAdsw. From φραγέλλιον --Ἰ,αῦ. flagelium by consonantal dis- 
similation (ὃ 42). Only NT and eccles. Siitterlin 120 curiously 
derives from °An. 


(2) Denominatives from -a- stems. 


These are ζημιόω (though Siitterlin 123 derives from ἔζήμιος on the 
ground that the fem. abstracts in -ἰὰ are probably all secondary forma- 
tions, see above, ὃ 136), ζυμόω, Tyopddw (Hellenistic, as also is peta’. 
tZup° is d.A. in Ph 310 (N°DcEKL), where συμμορφίζω is correct rdg. See 
below, ὃ 173), pildw (class., but Τέἐκριζόω LXX and NT only. ’Ek° is 
only inform acpd. See ὃ 119 and Vocab. s.v. Note the different senses 
in which the two verbs are factitive), σκηνόω (class., but ἐπι’ Hellenistic, 
kata Xen. and Hell.), Ἱτεφρόω (ἡ τέφρα ashes), Γφυσιόω (-- φῦσα bellows. 
NT and eccles. for class. “ιάω). 


(3) Denominatives from other stems. 
Stems ending in -es-. 


From nouns, σκοτόω (possibly from ὁ σκότος, then attributed to τὸ 
σκότος, hence by analogous formation), ἑλκόω, κυρόω, σθενόω (a word 
only known from 1 Pet 5’, and its mention in Hesych.), ὑψόω (though this 
may be influenced by its antonym ταπεινόω « ταπεινός). From 
adjectives, ἀκριβόω, mAnpdw (but Siitterlin 117 connects this with -o- 
stem, and cft. Lat. plerus). 

Consonantal stems, πυρόω, Txapitéw. 

From comparatives, ἑσσόω (2 Co 128 y*BD*, see above, p. 107, and 
Vocab. s.v. ἡττάομαι), ἐλαττόω. Debrunner (Worth. 103) traces these 
verbs in -ow back to the neut. ἔλαττον, which resembles the neut. of -o- 
adjectives. 

For téfoudevdw (Mk 912 AC), “Oevdw (ib. N69) see §§ 46, 119, and 
Vocab. s,v, 


§ 163-164] SUFFIXES. 397 


The dominant force of these verbs in -ow was instrumental or 
factitive, and as words tend to fall into groups under the stress 
of similarity of meaning, this kind of relationship was a deter- 
mining factor in the history of the growth of this class. 


Instrumental conception shown in various ways, €.g.— 


(a) To present, or reward with ; to injure, or punish with : 

Σημειόω, θανατόω, μαστιγόω, vexpdw, σταυρόω, TTaptapdw, τυφλόω, 
ἱφραγελλόω. Perhaps we may extend to κημόω, φιμόω, μισθόω, 
στεφανόω, Txapitdéw (fo endue with χάρις, see J. A. Robinson Hphes. 227). 


(b) To treat with kindness, or with eagerness, with evil, with guile ete. : 

Τδολιόω (LXX and NT. See Vocab. s.v.), Ἑηλόω, ζυμόω, κακόω. 

(c) To give expression to personal feelings : 

θυμόομαι, which may have influenced ἐμβριμόομαι (for the form see 
above, pp. 198-201; for the derivation, Debrunner in If xxi. 53; for 
this group of words, Siitterlin 125). 

The Factitive conception is evident in a number of groups, e.g.— 


(a) Words meaning to make strong ete. : 

tikavdw, Tkpatatdw (late form for “rive, see Vocab. 8.v. κραταιός), and 
ticyupdo (LXX, Herm. Mand. v. 28), from -o- stems, are accompanied by 
+Buvapdw (Vocab. s.v.), Tobevdw. 

The group, which is quite Hellenistic, may have originated in such 
pairs of contrasts as ἀσθενέω (to be weak): ἀσθενόω (to weaken—as early 
as Xen. Cyr. I. v. 3), καρτερέω : καρτερύω. On the other hand, the group 
may have started with the class. BeBatdw. 


(b) A contrast of meaning may have produced ὑψόω (from an -es- 
stem) to match ταπεινόω, and mAnpdw against κενόω. (See, however, 
above, under (3)). 


(c) A very important variety of the factitive meaning is found in a 
group of verbs in -de, derived from adjectives of moral as distinguished 
from physical meaning. Here the meaning is fo regard as, to treat as, not 
to make. Thus ἀξιόω (see Lightfoot Notes on Epp. 105), δικαιόω (see 
Evans ap. 1 Co 6" cited by 8 and H Romans 30). 


4. Verbs in -iw. 


§ 164. This small class of denominatives from :- 
stems was almost confined to Homer. There are no 
representatives in the NT, and the noun μῆνις which is found in Hermas 
(Mand. v. 33), has produced a verb μηνιάω (Sim. ix. 23%) according to 
the later formation. (See § 160 (4) above.) 

Two verbs ἐσθίω and κυλίω are later forms of ἔσθω and κυλίνδω, as 
shown on pp. 238, 246, 


΄ 
“ἰὼ 


398 A GRAMMAR OF NEW TESTAMENT GREEK. [§ 165-166 


5. Verbs in -ύω. 


§ 165. A small class of denominatives in v- stems 
is represented in the NT by ἀρτύω from ἀρτύς, which 
Hesych. gives as Ion. for ἀρθμός, cf. Lat. artus. (The same root as in ἄρτι, 
dpapicxw.) For the transition from original meaning to prepare to later 
meaning to season, see Vocab. s.v. and Wakernagel’s important note in his 
review of Mayser Gram. i. (ThLZ xxxiii. 36, n.1). Δακρύω, ἰσχύω, μεθύω 
«μέθυ, wine). 

Two other verbs κωλύω and μηνύω are of doubtful etymology (see 
Boisacq 8.vv.) ; κωλύω probably belongs to Class I. a. (a) above. 

In the following the suffix -yo is added to a root: 

Θύω (<*dhu-ijo), -πτύω (ἐκ-, éu-) (*[s]piz-1d, Lat. spuo), φύω (Lesb. 
prio «- Ἐφυιω), ῥύομαι (<*Fpu-, *yri-, see Boisacq 846). 

Two other stems are represented by compounds. Καμμύω, a syn- 
copated form of καταμύω, is a Kown word (see Ruth. NP 426f.). For 
ὠρύομαι see above, pp. 330, 387 n.1. Its root (I.E. *(e)réu-=“‘cry’’) is 
quite distinct from that of ῥύομαι above. 


, 
=UW 


6. Verbs in -εὐω. 


§ 166. The denominatives were first formed from the stem 
of nouns in -evs. Strict phonetic law would require a verb in 
-elw from *-nF-1- or -*eF-1-, but the influence of the nom. sing. 
-ev-s, and perhaps the analogy of other tenses (cf. καίω, *xaFio : 
éxavoa) determined the form -εὔω. Thus -aiw : -avoa: -atw: : 
-elW : -ευσα : -εύω. 

As the suffix -evs in a noun stands for one holding a pro- 
fessional rank or vocation, so the corresponding verb in -evw 
marks the exercise of that profession. Thus βασιλεύω, βραβεύω, 
épunvedw, ἀγρεύω, ᾿ἁλιεύω. This type became very productive 
at an early stage of the language and spread to other stems than 
nouns in -evs, largely under the influence of semantic analogy. 
No doubt the relation ἵππος : ἱππεύς : ἱππεύειν partly accounts 
for the freedom with which this suffix was added to stems in -os. 


The following groups are specially noteworthy :— 

(a) Domination, oversight, rank : 

Βασιλεύω (which βραβεύω regularly follows) gives by example ἡγεμονεύω 
(μων), TOprapBedw, καταδυναστεύω, κελεύω, κυριεύω (kata), παιδεύω, 
πρεσβεύω, ἐπιτροπεύω,  ἀνθυπατεύω (Ac 1813 HLPSs, Mart. Polyc. 21), 
ἀγγαρεύω (for this “* interesting old Persian word,” and the noun ἄγγαρος in 
Aesch., see Vocab. s.v.). So perhaps ἐποπτεύω ( - της, originally 
overseer ). 


§ 166] SUFFIXES. 399 


(6) Correlative to these are verbs for subjection, obedience and service; 
especially religious service, thus leading on to verbs which mark the 
exercise of religious functions : 

Δουλεύω, λατρεύω, νηστεύω, Τἱερατεύω (see Hort Comm. 1 Pet. 109), 
μαντεύομαι, Tpeotevw, παρεδρεύω (1 Co 913, “ kultischer Terminus,” 
Lietzmann HNT in loc. See also Vocab. s.v. Upoo is read by KL and 
later MSS), προφητεύω. For μαθητεύω see note below. 


(c) Verbs connected with hunting, snaring, etc. : 

᾿Αγρεύω, θηρεύω, ἐνεδρεύω (ὃ 118), Τἁλιεύω, ᾿παγιδεύω and possibly 
also φονεύω and τὀλοθρεύω (late form of ὀλεθρεύω). Κατατοξεύω (Heb 
1220 (LXX)) finds its way into T.R. on the strength of “‘ nonnulli minuse.”’ 


(4) A number of these verbs mean “‘ to play the part of,” “το act as” 
the man denoted by the corresponding noun : 

Καπηλεύω (dos), to play the huckster, to deal in for purposes of gain. 
For pap. illustrations in support of RVmg., see Vocab. s.v. So μαγεύω 
to play the magus, deal in magic ; πολιτεύω, in LX X and NT always ‘Tevopat, 
0 act the citizen, live one’s public life; πυκτεύω (rns) act like a bower, fight ; 
Tyupvitedw (spelling, p. 72), to be scantily clad (as in Dio Chrys.; but in 
Plut. and Dio Cass. to be light-armed, where we may possibly trace a noun 
γυμνίτης tormed after ὁπλίτης ; see Brugmann Gir.t 237). The meaning 
“to behave as”’ is clear in μοιχεύω (χος), πορνεύω (vos). 


Many of the verbs in -evw fall naturally into none of these groups. 
They are either factitive, or connote the possession of a quality, or re- 
present an action. In some instances we find the relationship of quality 
passing to that of action. 

The factitive meaning is evident in— 

Δεσμεύω to bind (“yds), σαλεύω to shake (dos a trembling), Ἰσωρεύω 
(°pos a heap) to heap on, overwhelm, φυτεύω to plant, Τἀποδεκατεύω (see 
“dw above, ὃ 163). 

The possession of a quality is conspicuous in— 

᾿Αληθεύω to be truthful, hence to deal truly (Field, Notes 192, Vocab. s.v.), 
εἰρηνεύω, ζηλεύω (late and rare for -dw), μνημονεύω, περισσεύω 
(passes from the intr. to the trans., see Lightfoot Notes 48, Milligan Thess. 
44), πιστεύω (passes from the intr. to the trans., but in the NT the 
activity of faith is strongly present in the word), πτωχεύω, συμβουλεύω 
(<°hos). The cpd. φιλοπρωτεύω, ἁ.λ. in 3 719, is from φιλόπρωτος 
(Polyb., Plut., Artem.), the simplex πρωτεύω (LXX, Col 118) is later class. 
(Plat., the Orators etc.). 

The idea of action appears in the following : 

᾿Αγορεύω to speak in the ayopa (contr. ‘palw, ὃ 172 (d)) lost its specific 
meaning, and its cpd. mpooayopedw=to address, hail (so Heb 51°, see 
Moffatt ICC in loc.). See Vocab. s.v., and for use of ἀγορεύω and cpds., 
Rutherford NP 326 ff. Avavuxtepedw, a Hellenistic coinage (Xen.), 
following διημερεύω which occurs once in Plato, afterwards in Hellen. 


400 A GRAMMAR OF NEW TESTAMENT GREEK. _ [§ 166 


authors. Ἐμβατεύω (see ὃ 118 (b), Vocab. s.v.), Τἐπυγαμβρεύω (see ὃ 120, 
Vocab. s.v. γαμβρός), κινδυνεύω, κυκλεύω (rather rare word for common 
“έω, “ow, Jn 1074 (B), Rev 20°, see WH App.? 178, also Vocab. s.v.), 
μνηστεύω (as old as Homer, «μνηστήρ <pvdopa. See Boisacq 641), 
Πορεύω in class. Gk. was used with a transitive force, but only the far 
More common “evopat survives (with its 8 epds.) in the NT. Its 
synonym δὁδεύω was an early poet. word which reappeared in Hellen. 
prose. Mefedw (poet. until Xen.). The cpd. faixpadwredw (<°ros, 
§ 105) probably followed the analogy of φυγαδεύω (Debrunner in Blass 
Gr. 65). Lobeck (Phryn. 442) observes “‘ Extrema Graeciae senectus 
novum palmitem promisit αἰχμαλωτεύειν.᾽" 

Although the active ending -ev could, and did originally, convey the 
dea of “‘ being or behaving, or acting as ἢ the influence of other 
classes of verbs where this condition is expressed by the middle (e.g. 
ἄχθομαι, αἰδοῦμαι) led to the formation of deponents in - εύομαι : 

᾿Εγκρατεύομαι (< THs, § 118 (c)), ἐμπορεύομαι (<‘pos, assumed this 
form through mistaken connexion with πορεύομαι, see ἃ 118 (a)). 
}Neptepevouar (<°pos, ὃ 104) is first met with in 1 Co 184; elsewhere 
only in Marcus. Πραγματεύομαι (<mpadypara, business) to trade, a 
common commercial term in papp. See Vocab. s.v. Ἱπαραβολεύομαι 
(<-apdBodos, ὃ 123 (b), according to the true rdg. Phil 2°°, rather than 
trapaBouvrkevouac (CKLP), see Kennedy, ΜΟΊ in loc.). Τ'Ρυπαρεύομαι 
(read by 046 ap. Rev 2211, see under ῥυπόω above, ὃ 163). Στρατεύομαι 
(στρατὸς, army in camp) fo serve as a soldier, make war. Depon. only is 
found in NT, though -εύω is occasionally used in class. authors. In later 
Hellen. active used transitively enlist. ΤΧρηστεύομαι (from “τός) not 
found earlier than 1 Co 134. 

Note.—In several of these verbs we observe the transition from intr. 
to trans. use : 

᾿Εμπορεύομαι to travel as a merchant (Jas 413), ‘ then, with a transitive 
force “‘ to import,” “‘ purchase,” “‘ traffic in,” ‘‘ make gain or business of,” ’ 
Mayor, ap. 2 Pet 23, This meaning of to exploit occurs in Polyb. xxxviii. 10, 
where the active form is used. Τθριαμβεύω to celebrate a triumph (so 
in Plut. eg. Tib. Grac. 21, C. Grac. 17). Then, as in NT, to lead in 
triumph. So Col 215, 2 Co 214 (on which see Lietzmann HNT' in loc., 
with parallels from Plut., also Vocab. s.v.). +Ma@ytevw, intr. to be a 
disciple (Mt 27°’ ABL. So Plut.), and trans.=to make a disciple (Mt 
2819, Ac 141); -εύομαι, depon.=to be a disciple, (Mt 1352 2757 NCD, 
1 55. 1) 


99 66 


11. Consonantal yo- presents. 
These consist of A. -n-do-, (1) αίνω, (2) -ύνω. 
B. -γ-ἴο-. -[-ἴο, (1) -aipw, -elpw, -ipw, (2) -ἄάλλω, -έλλω, 
-iAXw, -ύλλω. C. -πτω, -cow. D. -ζω, (1) -dlw, (2) -ίζω, 
(3) -ύξω, (4) -ἔξω, (5) -όζω. 


§ 167] ες SUFFIXES. 401 


_ A. Verbs in -n-10-. 
1. -αίνω. 
§ 167. These denominatives were formed 
from a variety of stems. Starting with stems 


containing, or ending in, -n-, the suffix was extended under the 
influence of analogy to other groups of words. 


, 
“αινῳω 


Thus : 

(a) To stems with -n-: Ποιμαίνω (from ποιμεν- <*-mn-i), εὐφραίνω 
(from εὐφρον-), βασκαίνω (fr. βάσκανος <*Bax-oK-avo-s); κερδαίνω (present 
stem not found in NT) may come from *xépdav (>’Axépdev prop. name, 
also Lat. cerdo, -dn-is), see IF xxi. 20. 

This applies also to neuters in -μα (<*-mn-; cf. ὄνομα, Lat. nomen): 
Σημαίνω (from σῆμα <I.E.* dhid-mn), θερμαίνω (probably from a neut. 
Ἐθέρμα), and Aupaivopat (from λῦμα, ποῦ λύμη, see IF xxi. 22). 

(6) The suffix is often added to stems in -pos. Debrunner (following 
Brugmann Grd.? τι. i. 347, 578) recalls the interchange of the n- and r- 
suffixes in J.E., and thus accounts for the partiality of the ro- and lo- 
stems for -αίνω (JF xxi. 31). 

Μωραίνω from “pds, ξηραίνω from “pds, πικραίνω from “pds, illustrate 
this partiality, whilst pratvw from μιαρός is an example of the inter- 
change of n- and r-. 


(c) Certain verbs are treated by Brugmann (G7.* 349) as instances of. 
the zo extension of a nasal present. To this class belong : 

Μαραίνω (cf. Skt. mr-na-ti) <I.E. *mera”- (grind), *mere- (die) identical 
in Lat. morior, βροτός ; mer-n-i0 >papaive (Boisacq). 

Ὑφαίνω <*uébh-p-jo >id-avio >idaive. 

(d) The analogic spread of this suffix in verbal formation was aided 
by certain groups of words. We need only consider the bearing of this 
upon NT vocabulary. In the factitive group μελαίνω (from μελαν-) 
quite naturally led to λευκαίνω from λευκός, whilst θερμαίνω set the 
fashion for a whole group of words of which ξηραίνω is our only example. 
Under the influence of μιαίνω the Homeric ῥυπόω gave place to ῥυπαίνω 
(Xen., Arist.), with the result that an adj. ῥυπαρός was formed on the 
reverse analogy of μιαρός : μιαίνω.ς One may hazard the suggestion 
that πικραίνω was influenced by γλυκαίνω, which Brugmann (IF xxxviii. 
125 f.) has explained as a substitute for γλυκύνω by the principle of 
‘‘ prohibitive dissimilation.”” Τλυκαίνω is the one verb from a -v stem 
which joins the -aive class; contrast its synonym ἡδύνω. In the in- 
transitive group there are a number of words representing a state of mind 
or character, to which, in the class. age, pwpatvw belonged (see Note 
below), and others representing a bodily condition, e.g. ὑγιαίνω from 


ὑγιής. 


402 A GRAMMAR OF NEW TESTAMENT GREEK. [§ 167-169 


NotE.—As was the case with -evw verbs, we see a change from active 
to middle, and from trans. to intr., in the following: 

Μωραίνω (1) class. to be foolish. (2) In LXX and NT, factitive, to make 
foolish (1 Co 1*°), -opat, to become foolish (Rom 133), to become tasteless (Mt 
538 || Lk 1454). 

Πικραίνω to make bitter (Rev 10°), -opat to show bitterness (Col 319). 


2. -vvo. 

ὃ 168. Verbs in -ύνω stand in close relation 
to adjectives and nouns with v- stems. But 
since some adjectives in -vs were related to neut. nouns in -os, 
the verbal suffix -vvw came to be attached to other stems in -os 
where no such adjectives in -vs were to be found. 


, 
“υνω 


(a) In ἀμύνω and πλύνω we have primary verbs, where -ὖν- is part of 
the stem (see Debrunner /F xxi. 73). 


(b) Adjectives in -ύς furnish us with βαθύνω, βαρύνω, βραδύνω, 
εὐθύνω, (παρ-)ὀξύνω, παχύνω, πλατύνω ; and πληθύνω may have been 
formed from the Homeric noun πληθύς. 


(c) Αἰσχύνω comes from αἶσχος, which had no corresponding adj. 
in -vs, (αἰσχύνη being a back-formation from the verb). 


(4) Βαθύνω and πλατύνω became patterns for a group of words 
represented by peyadtvw, μηκύνω, and possibly πληθύνω. Παχύνω prob- 
ably led the way for σκληρύνω (from σκληρός). 

(6) Μολύνω is closely connected with μέλας (<*uéAavo-s), cf. Skt. 
malind-s, 1.E. mejno-s. ᾿Οτρύνω (παρ-) is really a compd., <*6-rpu-v-tw 
«ὁ (see § 111 above, and Brugmann Gird.? τι. ii. 817), and rpu (ef. Skt. 
tvar-, Brugmann ἐδ. i. 260, 311). 

(f) Other -n-yo- verbs: Kpivw (xpi-v-1w), ἐλαύνω (see p. 235. Solmsen, 
Gr. Wortf, 51 -- ἐλα-υν-τω « ἐλα- ἔων, nomen agentis from ἐλα- in ἐλά-σαι, 
ἐλα-τήρ ete.). 


B. Verbs in -r-qo, -l-20-. 


δ 169. (1) In -aipw we have two primary verbs 


aii αἴρω and χαίρω, and the denominative καθαίρω (from 
καθαρός). 
; In -eipw almost all the verbs are primary, thus, 
-εἴρω 


ἐγείρω, κείρω, σπείρω, φθείρω. 

For ὀμείρομαι see p. 251 above. 

Οἰκτείρω is another spelling for οἰκτίρω, which Brugmann Gird.* τι. i, 358 
derives from *oixri-po- (cf. οἰκτίζω) after the type ὀλοφύ-ρο-μαι. 


§ 169-171] SUFFIXES. 403 


In -vp@ we have two primary verbs, πτύρομαι 
and σύρω, and the denominative μαρτύρομαι. 

(2) The only verbs in -ἄλλω in the NT are primary, 
viz. ἅλλομαι, βάλλω, θάλλω, ψάλλω. The same 
applies to verbs in -έλλω, viz. μέλλω, στέλλω, τέλλω, except ἀγγέλλω, 
which is a denominative from ἄγγελος. For ὀκέλλω see pp. 243 and 294. 

In -ίλλω and -vAAw we have only the primary verbs τίλλω and σκύλλω. 


-Upw 


-ddd\w, -έλλω 


C. Verbs in -πτω and -cow. 


1. §170. It is an open question whether any of 
the -rrw verbs (except denominatives as χαλέπτω 
«χαλεπός, none of which occur in the NT) were originally -io- verbs. 
The NT list consists of ἅπτω, βάπτω, βλάπτω, θάπτω, θρύπτω, καλύπτω, 


“πὴτω 


κάμπτω, κλέπτω, κόπτω, κρύπτω, κύπτω, νίπτω, ῥάπτω, {ἐπιράπτω, ῥίπτω, 
σκάπτω, σκέπτομαι, (ἐπισκ΄, late form οὗ ἐπισκοπέω), τύπτω. 

Of these βλάπτω and νίπτω, whose roots ended in a labiovelar (for 
βλάπτω cf. Skt. mare-, νίπτω </neig"-), must belong to the -7o- class, 
for *neig'-10> νίζω (cf. peq"id> πέσσω). Similarly with those whose roots 
ended in ¢, viz. βάπτω (ἐβάφην), θάπτω (/dhmbh-), κρύπτω (κρύφα), 
σκάπτω (ἐσκάφην, σκάφος), where the similarity of the aorist forms 
(ἔκρυψα : ἐκάλυψα) reacted on the form of the present. The remaining 
verbs in the list may quite well have come from stems in -i0. Σκέπτομαι 
may be <*oxemiopa «-ἕσπεκιομαι (=Lat. specio, Skt. pdsSyami). See 
Brugmann, Gr.* 343 f., also Debrunner JF xxi. 207 ff., Hirt Handb. 378 1. 


2. 8171. Verbs in -σσω are almost all from 


“σσω - 
guttural stems and most are primary. 


(a) Guttural stems.—Primary : Δράσσομαι (> δράγμα, δραχμή), -μάσσω 
(ἀπο-, ἐκ-), πράσσω (from *mpax(o)-, οἵ. mépa, mépa-v, Grd? τι. i. 
481), ταράσσω (rapay-), τάσσω (ἐ-τάγ-ην), φράσσω (*ppaxyw), πλήσσω 
(ἤπλάκιω, οἵ. πληγή), ῥήσσω (according to Wackernagel, Hellenistica 24, 
Fraenkel, Nom. Ag. ii. 40f.) supplanted ῥήγνυμι (to break) in the Hellen. 
age, under the influence of ἔπληξα: rAnoow. There was a distinct verb, 
Att. parte, Hellen. ῥάσσω Ion. ῥήσσω (to throw, strike, dash). Debrunner 
(Blass Gr.* 61) suggests that the two verbs coalesced in the Kown, and 
it is significant that the latter meaning fits the verb better in Mk 915 
(where, in fact, D reads ῥάσσει) and Lk 9%, also in the cpd. προσέρηξεν, 
Lk 6. For derivation of ῥάσσω see Vocab. s.v. ῥήγνυμι. Φρίσσω (cf. 
φρίξ), νύσσω, ὀρύσσω (Ἐὀρυχιω), πτύσσω (cf. πτύξ, -υχός. For etym. see 
Boisacq 824). Denominative: Πατάσσω (πάταγος), φυλάσσω (φύλαξ), 
χαράσσω (χάραξ), which is represented in NT by its derivative χάραγμα, 
ἀλλάσσω probably from the adv. ἀλλάξ (Debrunner IF xxi. 219), ἑλίσσω 
(see above, p. 236), βδελύσσω (βδελυρός, cf. Aesch. deA’Krporos), 


κηρύσσω (κῆρυξ). 


404 A GRAMMAR OF NEW TESTAMENT GREEK. [§ 171-172 


(6) Dental stems.—Primary: Πλάσσω (ἔπλαθιω, following the -x-. 
type in present, but aor. ἔπλασα, ἐπλάσθην. See Brugmann, (r.* 119, 
Anm.8). Denominative : πυρέσσω (πυρετός). 

(c) ᾿Εντυλίσσω was formed from the prepositional phrase ἐν τύλῳ under 
the influence of ἑλίσσω (see Debrunner IF xxi. 235, Wortb. 115); for its 
rarity and meaning see Abbott JV 346f. and Vocab. s.v. By back- 
formation a simplex τυλίσσω was made (see LS 8.v.). 

(d) Σπαράσσω (<I.E. *spereg-)is taken by Debrunner (JF xxi. 224) as 
a -¢w verb, which follows the analogy of ταράσσω in the present (Aor. 
also é€omdpaga); τινάσσω (ἄπο-, ἐκ-) of uncertain etymology; φρυάσσω 
(like its cognate φριμάσσομαι) a lengthened form, related to φρέαρ 
(<*ppyFap <I.E. *bhréup-); αἰνίσσομαι (>aivtypa) from αἶνος. The 
origin of the -ίσσομαι is unexplained. 


1). Verbs in -ζω. 
1. -ἄξω. 

§ 172. This suffix originated in the attach- 
ment of -c@ to stems in -a6-, thus λιθάς (λεθαδ-) 
λιθάζω. In a few instances a guttural stem was so used, e.g. 
ἅρπαξ (ἁρπαγ-): ἁρπάζω. The use was extended to neuter nouns 
with stems in -at-, 6.4. ὄνομα : ὀνομάζω, and then to nouns in 
-d- stems, 6.6. ἀγορά : dyopdétw. In time -ἀζω was a suffix that 
could be added to any stem, though it is generally easy to 
trace the influence of semantic analogy. 


-alw 


(a) How productive this suffix became may be judged from the fact 
that out of about 70 verbs in -ἀζω in the NT only two are derived from 
Stems in -ad-, the second being φράξω (<dpad-i1w, see Brugmann Gird. τι. 111. 
182; cf. φραδή, api-ppad-js). Σεβάζομαι probably came from σέβας, 
which has not stem in -ad-, but see p. 258 above; and σκεπάζω (whence 
ὅσμα) from σκέπας, -aos. 


(Ὁ) Κράζξω and σφάξω are the only other verbs (cf. ἁρπάξω) from 
guttural stems, though a confusion of flexion is found sometimes in the 
conjugation. See Proleg. 56, ὃ 95 above, and Vocab. s.vv. ἁρπάζω, 
βαστάζω. 

(c) Like ὀνομάζω are θαυμάζω (θαῦμα), χειμάζω (χεῖμα), δελεάζω 
(δέλεαρ. -ατος). 

(4) From nouns in -@ we have ἀγοράζω (see Rutherford NP 214), 
ἀκμάζω, ἀλαλάζω, ἀναγκάζω, αὐγάζω (see Vocab. s.v.), βιάζω, δικάζω, 
δοξάζω (§ 118), ἐξουσιάζω, ἑορτάζω, ἐπηρεάζω (see ὃ 120), θηλάζω, 
κραυγάζω, πειράζω (poet. and late prose for “pdw), παρρησιάΐζομαι, 


$ 172] SUFFLXES. 405 


σεληνιάζομαι (Mt 424 1715. A “‘late and rare’ word. ΤῸ reff. given in 
lexicons add Vett. Val. 11329, and “σμός, 127° 9°), σκευάζω (mapa-), 
σκιάζω (ἐπι-, κατα-), σπουδάζω, -στεγαζω (ἀπο-), σχολάζω, τυρβάζω 
(Lk 1013 AP), χλευάζω (see Solmsen 246 η.1). 


(6) From nouns and adjectives in -(c)os, -()ον : ᾿Ενυπνιάζω, ἐργάζομαι, 
θορυβάξω, Τκαυ(σ)τηριάζω, κοπάξω (κόπος), σινιάξω (ἅ.λ. in Lk 223! for 
Attic σήθω ; probably both σίνιον and its verb belong to the colloquial 
Kown), ὑπωπιάζω (from ὑπώπιον, but tpuwtdlw from μύωψ, see ὃ 108 
above, and Mayor Jude and 2 Pet 95n."), χορτάζω (for history of word see 
Lightfoot ap. Phil 41%). 

Τ᾽ Αγιάξω (on this new word coined (or appropriated) by Jewish piety 
see Vocab. s.v.), ἀτιμάξζω, yupvdlw, δοκιμάζω, Τἐνταφιάζω (§ 118), 
ἑτοιμάξζω, ἡσυχάξω, νηπιάζω (1 Co 142°, elsewhere only Hipp., for 
νῆὴπιαχεύω (Hom.), νηπιάχω), Ἱπαρομοιάζω (Mt 2327, elsewhere only 
eccles. ; for this and for the simplex read by B, 1, see ὃ 123), Γπυρράζω 
(Mt 162), from πυρρός ; elsewhere only in Byzantine authors. In LXX 
and Philo πυρρίζω. The -a¢w form may possibly be due to the accom- 
panying verb), totuyvdfw (LXX, NT, late). To these should be added 
διστάζω from ἔδιστος, cf. Skt. dvisthah (uncertain), ἐξετάζω (whence also 
ἐτάζω (Wis 21°) and Τἀνετάζω) from éreos (<*ereFos) true, real: see 
Vocab. 42a, κολάζω from κόλος docked (<I.E. *gold-). 


(f) From adverbs: Διχάζω (diya), πλεονάζω (πλέον). 


(g) From verbs: ᾿Αμφιάζω (on the form and spelling, see p. 228). ‘‘ The 
back-formation ἀμφιέζω (from the aor. of -έννυμι) is an obvious first 
step towards ἀμφιάζω, which shows the influence of the large class of -ἀζω 
verbs” (Vocab. 28a). For derivation direct from ἀμφί see p. 68. 
-βιβαΐω, causative of Baive, mostly in cpds. (in NT only a epd. with 
ἀνα-, ἐμ-, ἐπι-, κατα-, προ-, συν-), from Ba- <,/g d- in Skt. jigdti, cf. 
Lac. 3 pl. βίβαντι and Hom. pte. βιβάς (=striding). Δαμάξω (according 
to Debrunner Wortb. 122 n.*) is a back-formation from ἐδάμασα the aor. 
of the older present stem δάμνημι. Στενάξζω, originally a frequentative 
of στένω. Φαντάξω from φαίνω. 


(h) A few verbs come under none of the above headings : 
᾿Ασπάζομαι is derived from ἀ <*p (cf. prep. ἐν, ὃ 118), and ,/seq¥, seen 
in the Hom. ἔννεπε (<*éevoeme) ἕσπετε (<*éev-orere), Lat. inseque. 
Βαστάζω (for meaning and flexion see Vocab. s.v.), cf. Lat. gero (<*geso), 
gesto, to carry. 
Nuotdfw and its cognate νύσταλος from I.E. *sneudh- <*sna, cf. Lat. 
nare. 
+Midtw (see pp. 69, 254) from I.E. *pi-s(e)-d-, ef. Skt. piddyati (*pi-zd-). 
See Boisacq, s.v. Ἐξ - 
ΝΟΤΕ.---Ἰ. Semantic analogy may account for a few groups of words 
in -ἀζω related by common meaning rather than by similarity of stem : 


406 A GRAMMAR OF NEW TESTAMENT GREEK. [§ 172-173 


Thus for utterance of sound, ἀλαλάζω, κράζω, κραυγάξω, στενάζω: 

for reviling, ridiculing, reproaching, ἀτιμάΐζω, xevalw, ἐπηρεάζω : 

for testing, judging, separating, δικάζω, διχάζω, δοκιμάζω, ἐξετάζω, 
πειράζω, σινιάζω : 

for numerical relationships, following μονάς : μονάζω (a verb found in 
Barn 4°, Herm. Sim. ix. 26%), δυάς : δυάζω etc., διχάζω, 
διστάζω, πλεονάζξω: 

for mark of age, ἀκμάΐω, νηπιάζω : 

for state of health: the related pair ὑγιαίνω : ὑγιάζω represented the 
intrans. and the factitive conceptions. Hence *vocaive : νοσάζω. 

The obsolescence of νοσαίνω led to the use of the passive -dfopa, to 
mark the possession of a disease. Hence Τσεληνιάζομαι from which 
by analogous formation Deissmann (LAH 251, #256) accounts for δαιμονι- 
άζομαι, in Wesseley Zauberp. (=P Par 574) 8% 3007, 

2. The termination -a¢ does not always carry a transitive meaning. 
This is seen specially when it is attached to -o- stems, for -όω here has 
a prescriptive right to the factitive meaning. Rutherford NP 284 says, 
‘“ Verbs in -a¢w from adjectives in -os are rare at the best, and though 
ἀτιμάζω, διπλασιάζω and one or two more bear a transitive meaning, the 
majority of such words are neuter.’ The famous crux βιάζομαι is dis- 
cussed fully in Vocab. s.v. 


2. -ifw. 


“it § 173. The origin and extended use of this 
23 most productive suffix is closely parallel to 


that of -afw. 


(a) From stems in -:d-. 

᾿Αγκαλίζομαι (ἐν-), the simplex poetical, from ἀγκαλίς, Homeric for 
ἀγκάλη (Lk 338). Τ᾿ Ἐναγκαλίζομαι (LX X and Plut.) may be an old poet. 
word which survived only in vernacular. Mt and Lk avoid it when used 
by Mk 985 1015, For meaning see Hapos. 1x. ii. 300. +BodiLw (Vocab. s.v.), 
ἐλπίζω, ἐρίζω, μερίζω, pari (for origin and meaning, see Rutherf. NP 
264; Field Notes 105), ῥιπίζω (see Hort ap. Jas 15), σφραγίζω, φροντίζω. 
So παίζω is from πάξις. Αὐλίζομαι might conceivably come from αὖλις 
(Hom.), but is generally derived from αὐλή. Kopilo from κομίδη 
(Brugmann Grd.? wm. iii. 231: cf. Skt. gama-h (<I.E. *komo-s) cami, 
Boisacq 489. For various meanings of the word see Vocab. s.v.). 


(b) Other stems with nominatives in -is follow this type, with the 
mixed declension of ἔρις (acc. ἔριν and epida) as a possible link (so 
Debrunner Wortb. 128). 

Κιθαρίζω, ὑβρίζω, χαρίζομαι, and, through similarity of declension 
(see p. 140), ἐπελεκίζω (-- πέλεκυς. See Lob. Phryn. 341). 


§ 173] SUFFIXES. 407 


(c) Guttural stems supply μαστίζω (Ac 22%5, Ep. and late prose for 
μαστιγόω), σαλπίζω, στηρίζω (for mixed conjug. see p. 259), φλογίζω 
(--φλόξ, though possibly belonging to the class (h) below). 


(4) From -a- stems come αὐλίζομαι, ὀργίζω, Τσμυρνίζω (Mk 1523, 
mingle with myrrh. Elsewhere only Diosc. be like myrrh. In very late 
writers, embalm), διυλίζω (both simplex and cpd. are late—from ὕλη, 
sediment), φημίζω (see p. 387 n.1), Τφυλακίζω. 


(6) From (a) nouns and (β) adjectives, in-os (-ov). (a) ᾿Αφρίζω, Τἀνεμίζω 

(elsewhere only schol. on Od. xii. 336. For class. dveuow. See Hort ap. 
Jas 15, also Mayor, who shows that James has a fondness for verbs in 
-ἰζω), βασανίζω (βάσανος, touch-stone), βυθίζω, tyapifw (see N. 2 below), 
εὐαγγελίζω (§ 106), Τεὐνουχίζω, rhe ea first occurs Heb 1088, but éx6° 
twice in Polyb.), θησαυρίζω, tipatitw (Vocab. s.v.), Τκατοπτρίζω 
(κάτοπτρον, mirror), kevtpiLw (Xen.), Τκολαφίζω (not found earlier than 
NT, from κόλαφος, vernac. for κόνδυλοι, knuckles, see Lobeck Phryn. 175), 
ἵκρυσταλλίζω (4A. Rev 211, λογίζομαι, μυρίζω (Ion. and Comed. 
also pap.), νομίζω, ξενίζω, -οικίζω (kat-), Jas 45 NBA (see N. 2 below), 
ἐἱμετοικίζω ( (<pérotxos, Arist.), ὁπλίζω, Τὀρθρίζω (LXX and Lk 21% for 
class. ὀρθρεύω. Hellen. acc. to Moeris, see Thumb Hellen. 123), ὁρίζω, 
ὁρκίζω (sound Attic in spite of Phrynichus, as Demosth. Fals. Leg. 278 
shows. See Rutherford VP 466f., Lob. Phryn. 560f.), ὁρμίζω (προσ-), 
πλουτίζω, ποντίζω (κακα-), πορίζω (whence Τ᾿ σμός), τς, ῥαβδίζω, 
Ἰσκανδαλίζω (only LXX and NT, see Thumb Hellen. 123, Helbing Gr. 
127), σκορπίζω (an Ionic word, found in a fragment of Hecatzus, 
elsewhere only in Hellen. writers for Att. σκεδάννυμι, see p. 387 n.t. 
Rutherford NP 295: from σκορπίος, an engine for throwing missiles, 
hence to scatter), σπλαγχνίζομαι (Thumb ib., Helbing 7b., Vocab. s.v.), 
τραχηλίζω (Xen.; for meaning see Moffatt JCC ap. Heb 433), τυμπανίζω 
(Eupolis and Hellen., originally to beat a drum, τύμπανον, later to beat to 
death, see Moffatt ap. Heb 1135), φορτίζω, xpovilo, ψηφίζω, ψωμίζω. 

(8) “Ayvifw, ἀθροίζω (from ἀθρόος, assembled in a crowd, a copulative 
(see ὃ 107) for 4 under the law forbidding successive aspirates, and θρόος, 
noise<I.E. *dhreu-, *dhri-), (etxpokertta)( from °ros, see ὃ 105), ἀρτίζω 
([ἐξ- κατ-), Τἐνορκίζω (§ 118), ἐξυπνίζω (§ 119), ἰσχυρίζομαι (δι-, § 116), 
txabapi£w (for class. καθαίρω, see Vocab. s.v.), ee and ἜΤ 
(see Vocab. s.vv.), κουφίζω, μετεωρίζομαι (§ 122 (b), Vocab. s.v.), ὀρφανίζω 
(ἀπ-), προχειρίζω (§ 125), σοφίζω, ἐδ tera (Ph 9:0 SABD: 99: 
nowhere earlier), σῴζω, Τσυνετίζω (fr. συνετός), Herm. Mand. iv. 21, Ep. 
Diogn. 12°, (first found in Arist.). 


(f) From other 3rd Declension nouns and adjectives. 

Stems in -es- supply γεμίζω (unless this belongs to class (h) below), 
δανείζω (for spelling, see p. 77), ἐθίζω, Tédapitw (Arist. For meaning 
see Field Notes 74), θερίζω, ὀνειδίζω, Τσκοτίζω. ᾿Δλίζω (συν-) (--ἁλήν, 


408 A GRAMMAR OF NEW TESTAMENT GREEK. [$ 173 


a-Fadns, cf. ἀολλής : see Solmsen Gr. Wortf. 20), Τἀσφαλίζω (see Vocab. 
s.v.), ἀτενίζω, ἀφανίζω, and ἐμφανίζω (see pp. 236, 307). 

Stems in -ov-, -av-. Αγωνίζομαι, ϑαιμονίζομαι, ᾿κλυδωνίζομαι (Eph 
414, Elsewhere only Is 572°, Joseph., Vett. Val.), σωφρονίζω. 

Stems in -par-. ΤΔειγματίζω (Mt 119 8*BZ 1, Col 21%. See p. 320 η.3. 
For other exx. of this rare word see Vocab. s.v.), TSoypatilw (see 
Vocab. s.v.), Tkaupatilw, μετασχηματίξζω and συνσχηματίζω, τραυματίζω. 
Χρηματίζω covers two entirely distinct words: (a) to be called, «- χρήματα 
(<ypOpa, <*ypn-sopar) business, hence to do business under the name of 
X, to bear the name of ; (Ὁ) to warn, « χρῆμα (- χρῶ. « ἘἜχρη-(ω) Ξε χρησμός 
(<xp7¢{), oracle. 

Stem in-or-. Φωτίζω. 

Various stems supply Τἀλίξω (from ἅλς, ἁλύς : Mk 949, Mt 543, Ignat, 
Magn. 103, LXX ; first in Arist.), ἀνδρίζω, μακαρίζω (<pdxap), μυκτηρίζω 
(<puxrnp, nose), to sneer at, mock (see Vocab. s.v.), διαχειρίζω (δ 116 (a)). 


(g) From adverbs: Λακτίζω (<Adé, with the foot), to kick. Νοσφίζω 
(<vocdu, apart, aside) has in the two NT occurrences (Ac 5?-3, Tit 210) a 
special middle force, to purloin, supported by the papyri (see Vocab. s.v.). 
This poet. word first appeared in prose in Xen. Cyr. Iv. ii. 42, and is 
frequently found in Hellen. authors (as Wetstein shows). Xwpilw (<yepis)5 
Τἐγγίζω (- ἐγγύς) not very common in papyri (see Vocab. s.v.). 


(h) From verbs: Γνωρίζω from the same verbal root as γι-γνώ-σκω, but 
with the 7 that appears also in γνώριμος, norma (<*gnérima), then gnarus 
(<*@n-), narré (<*gnaré), ign6érd (Boisacq s.v.). The original causative 
force is largely lost, but is to be recognised in all its NT occurrences, even 
Phil 122 (see Vocab. s.v.). ᾿Ἐρεθίζω (ἐρέθω), ἱπροσοχθίζω (Heb 32° (xx) 
late form, rare outside the LXX, for προσοχθέω), πρίζω (pio); xpyto, 
to need, is closely related to ypaw, but both are derived from an old noun 
χρή (see § 107 under ἀχρεῖος ; also Boisacq p. 1069). φλογίζω (if from 
φλέγω, but possibly it comes from φλόξ, and belongs to class (f) above). 

The most common type of verbal derivative, as the ending -τίζω 
shows, is formed from the verbal adj. in -ros. These verbs are generally 
intensive or iterative. NT examples are αἱρετίζω (aipéoua, “perds) 
(Hipp. and inscrr. Polybius does not use the verb, but its deriv. 
°ruotns=partisan), βαπτίζω (Barre, Barres), Τῤῥαντίζω (ῥαίνω, pavrds). 


(t) The productivity of this suffix may be seen from the freedom with 
which new verbs were coined from prepositional phrases. This began in 
the class. period, 6.5. σκορακίζω -εἐς κόρακας (βάλλω), but grew consider- 
ably in later Greek. In the NT we have ἀποστοματίζω (see ὃ 115, Vocab. 
s.v.), but ἐπιστομίζω (see § 120, for form cf. ἐνστομίζω in P Par 5742174), 
both of classical origin, fdmokepadiLw (see 88 108, 109), Tévwtifopat 
(§ 118), Τἐξουδενίζξω (Lk 2311 W, see § 119), which are all Hellen., 
κατακρημνίζω (see §§ 109, 121) first appears in Xen, 


§ 173] SUFFIXES. 409 


As an example of new formations in papp. οἵ. ἐπιβωμίζω PSI iv. 4358 
(258 B.c.). 


(2) A few root verbs end in -ἰζω and may be given here: 


ἵζω (καθίζω) <*si-zd-6 or *s,d-jo < ,/sed, cf. ἔζομαι. 

κτίζω, cf. Skt. kséti, ksiydti (see Boisacq s.v.). 

σχίζω, cf. Skt. chid-, chindtti, chintté, to cut, split, Lat. scindo, pf. scidi 
(see Boisacq 8s.v.). 

τρίζω (Mk 918 zp. τ. 66. to grind the teeth), a word found from Homer 
onwards for the utterance of any sharp sound, from I.E. 
*(s)trei-g- with *stret-d- in Lat. stridéo. 

Notr.—1. This suffix was freely used in coining words on the analogy 
of groups with similar meaning. Thus (a) ᾿Ιουδαΐξω follows the well- 
known type of “ imitatives,” μηδίζω, λακωνίζω and even φιλιππίζω, to 
ape the Mede, to imitate the Spartan manners, to side with Philip, to which 
we may add ἑλληνίζω (> ᾿ἱστής) to Hellenise, to speak Greek. (Ὁ) The 
suffix was commonly used for the celebration of a festival ; thus σαββατίζω 
(LXX, Logion in P Oxy i. 1, σμός Heb 4°) follows the example of πανη- 
γυρίζω (<tmanyupts). In this way may have arisen γαμίξω (on which see 
N. 2 below), after the type παννυχίζω. (c) The only other group that calls 
for mention here is that of verbs describing a sound, whether vocal or 
instrumental. In the NT κιθαρίζω, σαλπίζω, tupmavitw, originated 
thus : see (c) above. 

2. The meaning of a verb in -ίζω often depends on the context, as 
Rutherford observes (VP 179). Sometimes the instrumental force is 
obvious, as in ῥαβδίζω, or the factitive, as in γεμίζω, or the causative, as: 
in ποτίζω ; but the lists given above show with how many verbs such 
a ready decision is impossible. Consistency in verb formation was some- 
times hindered by the previous appropriation of a suffix to another root. 
An instance of this occurs in 1 Co 147, where αὐλέω and κιθαρίζω are 
parallel. Here the verb expected, αὐλίζω, had been coined from the root 
αὐλή as early as Homer, with an entirely different meaning (see (a) and 
(d) above). An important point arises over the distinction of meaning 
between a verb in -ἰζω and a cognate formation. Sometimes the primary 
distinction between intrans. and trans. is maintained, as in σωφρονέω and 
owhpovitw, πλουτέω and πλουτίζω. On the other hand, although καθέζομαι 
and κάθημαι were available for the intrans. sense, καθίζω is intrans. in 
more than 20 passages and causal in only 3, unless we add Jn 198 as evi- 
dently interpreted by Justin M. (Ap. 1. 35) and Ev. Petr. iii. 7. The 
compounds of καθίζω are all used intransitively (ἀνα- Lk! Act, ém- 
Mt}, mept- Lk!), with the exception of συν-, which is intrans. Lk 2255 RAB, 
but trans. Eph 2°. The meaning of yapifw in 1 Co 738 has an important 
bearing upon exegesis. The verb is not found outside the NT, and here only 
in the active. It has generally been assumed that γαμίζω must stand 
in causative relation to yapew, but apart from exx. given above, we have 
the pairs torepéw : ὑστερίζω, κομέω : κομίζω, to remind us that this 


410 A GRAMMAR OF NEW TESTAMENT GREEK [8 173-174 


distinction is not invariably observed. Lietzmann HNT' in loc. cites 
χρονίζω, ἐλπίζω, ἐρίζω, ὑβρίζω as intrans. verbs with this suffix, and calls 
attention to the tendency in Hellenistic towards the intrans. use of trans. 
verbs (so Radermacher Gr. 18 f. 222 f.). He follows Wendland in conjectur- 
ing that itacistic pronunciation, ἐγάμησατεἐγάμισα, may have led to the 
confusion. It is significant that in the only other passages where the 
word occurs, Mk 1225 (and |{s), Lk 1727, γαμίζεσθαι means no more than 
γαμεῖσθαι. 


§ 174. This suffix was mostly used in onomatopesic 
formations, as γογγύζω (papp., Epict., Marcus, as well 
as LXX and NT; classed as Ion., not Att., by Phrynichus ; see NP 463 
and Vocab. s.v.), ὀλολύζω (see Vocab. s.v.), κατακλύζω from κλύδων 
(common in papp., Vocab. s.v.). 


4. -ἐζω. 
᾿Αμφιέζω (see pp. 228, 294, and 405 above). 
Καθέζομαι (from /*sed). Πιέζω (see above, p. 254 


-ἐΐω 


and Vocab. s.v.). 


The primary verb ὄξω (cf. ὀδ- μή), and the denomina- 
tive ἁρμόξω (cf. ἁρμόδ-ιος) the Hellen. spelling for 
Att. -rrw ; see Lobeck Phryn. 241, and, for other reff., Vocab. s.v. 


APPENDIX 


SEMITISMS IN THE NEW TESTAMENT 


VOL. II. PART I11.—25 


SYLLABUS 


+ 
a  - 
Ω 


PAGE 
INTRODUCTORY : 5 : 2 : : π΄ 215 


A. GENERAL STYLE AND STRUCTURE OF SENTENCE Σ 410 


1. Position of Verb in Sentence. 

2. Parallelism. 

3. Tautology. 

4. Parataxis. 

5. Casus pendens followed by Resumptive Pronoun. 

6. Constructions with καὶ ἐγένετο. 

7. Co-ordination of Participle or Infinitive with Finite Verb. 


B. Tuer SEVERAL PARTS OF SPEECH. : Ε : . 430 


1. Pronouns. (a) Definite Article: (δ) Personal Pro- 
nouns: (c) Reflexive: (d) Indefinite: (e) Relative: 
(f) Distributive Pronouns and Pronominal Adjectives. 

2. Numerals and Distributives. 

3. Adjectives and Adjectival Substitutes. 

4. Degrees of Comparison. 

5. Adverbs and Adverbial Locutions. 

6. Verbs: (i) Voice; (ii) Infinitive; (iii) Participle ; 
(iv) Indicative. 

7. Nouns: Cases. 

8. Prepositions and Improper Prepositions. 

9. Conjunctions and Particles. 


C. MISTRANSLATION OF SEmMITIC WORDS OR PHRASES : . 470 
lp ΜΠ Be VRE Νὺ OREO οδ΄ 11. IE ORS 7751 85 5159 2155 
Ac 947 316 429 1128, 
Ὁ ness 856 Rev 101 193: 2* 155-555. 


1). GENERAL SUMMARY. ‘ : 5 : 2 - All 


1. Classification of Semitisms. 

2. The Semitic Stamp in Translation Greek. 

3. Conditions under which the Gospel Tradition became 
fixed. 

4. The Several Books. 


SEMITISMS IN THE NEW TESTAMENT 


THE right of such an essay to a place in a volume dealing with Accidence 
and Word-formation is obviously open to challenge. It may be defended 
on two grounds. First, it will be highly convenient when such con- 
structions call for treatment in the volume on Syntax to be able to refer 
to their systematic classification in the present context, and thus to 
save space where compression is most needed. Secondly, the student 
who has found multitudinous references to this subject scattered through 
every part of the Prolegomena, both in the main text and in numerous 
additional notes, and who has felt the inadequacy of the brief survey in 
the Introduction to this volume, will rightly expect a more detailed and 
methodical investigation at the earliest possible place in this Grammar. 
To the Editor there is the additional and sufficient reason that such was 
the design of Dr. Moulton himself. 

Before entering upon this survey, however, it seems necessary to 
remind the reader that in some respects Dr. Moulton’s attitude to the 
subject of Semitisms in the New Testament was slightly modified after 
the first edition of the Prolegomena appeared. His main concern was to 
support Deissmann in his contention that the New Testament was written 
in no Judaeo-Greek jargon but in the lingua franca of the first century. 
He would have quoted, with some exegetical freedom, the saying of a 
second-century writer, ““ The Christians use no strange variety of dialect.” * 
But while he maintained this thesis to the end, a comparative study of 
the successive editions of the Prolegomena, of the articles in Cambridge 
Biblical Essays and Peake’s Commentary on the Bible, and of the Intro- 
duction to the present volume of the Grammar, will reveal a progressive 
tendency to do full justice to the influence of translation where Semitic 
originals may be posited with good reason. It must be remembered that 
Wellhausen’s Hinleitung in die drei ersten Evangelien only came into the 
author’s hands when the proofs of Prolegomena were in an advanced 
stage.2 But the cumulative evidence set forth in that book, supported 
by Syriac parallels to which his attention was drawn by Dr. Rendel 


1Hp. ad Diognetum, 5*: οὔτε γάρ που πόλεις ἰδίας κατοικοῦσιν οὔτε 
διαλέκτῳ τινὶ παρηλλαγμένῃ χρῶνται οὔτε βίον παράσημον ἀσκοῦσιν. 
2 See Ῥγχοῖορ." xii. f£. Additional notes were inserted at once at pp. 4 n.%, 
11 π.1, 14 n.3, 58 n.1, 91, 94 π.1, 97 also n.4, 139 n.1, 163 n.?, 213, 224, 226° 
231, 233, 235, 236%, 237, 240, 241 ἢ, In the 2nd and 3rd editions further 
points were considered on pp. 244, 247, 249. 
413 


414 A GRAMMAR OF NEW TESTAMENT GREEK. 


Harris, unquestionably restrained the ardour of the “ grammatical 
anti-Semitism ” with which Dr. Moulton has so often been charged.* 
Some years later, after reading Studies in the Apocalypse, he wrote with 
reference to the grammatical anomalies of the Book of Revelation, “‘ Dr. 
R. H. Charles has recently shown how many of its mannerisms are due 
to a literal transference of Semitic idioms.’ None the less Dr. Moulton 
insisted that many locutions which suggest Semitic idiom to the specialist 
in Hebrew or Aramaic fall within the range of late vernacular Greek, and 
he was careful to distinguish between pure Semitisms and those to which 
that title can only be given in a secondary sense.? The importance of 
these ‘‘ secondary Semitisms” is best seen when we “ are seeking for 
evidences of Semitic birth in a writer whose Greek betrays deficient 
knowledge of the resources of the language.’ A subtler test than that 
of pure Semitisms is found “in the over-wse of locutions which 
can be defended as good Kowy Greek, but have their motive clearly 
in their coincidence with locutions of the writer’s native tongue.” 4 
This statement of the case satisfies two such able critics of 
‘“ Deissmannism’’ as Professor G. C. Richards® and the late Canon 
C. F. Burney.® 

One further point must be mentioned, though obviously this Appendix 
is not the place for its full discussion. In an additional note to the second 
edition of Prolegomena,? Dr. Moulton hinted at the possibility that 
Aramaic-speaking populations in Egypt may have infected the Κοινή of 
that country. In the preface to the third edition® this matter was 
examined in reply to criticisms offered along this very line by Drs. Redpath, 
Swete and Nestle. Dr. Moulton’s fullest answer, supported by the 
weighty authority of Dr. A. S. Hunt, is to be found in Cambridge Biblical 
Essays. Quite recently the argument has reappeared in two forms. 
Canon Burney has attempted to turn Deissmann’s flank by quoting 
the word μαγδωλοφύλαξ from the very papyrus letter written by two pig 
merchants at Arsinoe which Deissmann had used as an example of the 
paratactic style of sources where no Semitic influence can. be predicated. 
Now this word contains the same root as the Hebrew δ, and is un- 


questionably Semitic, as Drs. Grenfell and Hunt show in their notes 


1 Pere Lagrange has phrased it happily: “Il n’en est pas moins vrai 
que lorsqu’ un helléniste ouvre le NT, en particulier les évangiles, il se trouve 
transporté dans les tentes de Sem. L’exagération de quelques hellénistes 
a été, reconnaissant chaque objet comme déja vu dans le domaine de Japhet, 
de prétendre quil en venait toujours” (S. Luc, p. xevi). 

2 Peake’s Commentary on the Bible, 592 ὃ (art. ‘‘ The Language of the New 
Test.’’). 

3 Supra 14-18. This should be borne in mind in qualifying two footnotes 
by Dr. Charles (JCC) Revelation, i. pp. x, n.*, exliii,n.. Vide supra, 33 f. 


' CBE 414. 5 JTS xxi. 286. 
6. Aramaic Origin of Fourth Gospel, 7. 
ΤΡ, 242. ® Pp. xvi ff. 


® Pp. 468 ff. 10 Aram. Orig. 5 1. 


SEMITISMS IN THE NEW TESTAMENT. 415 


on the document.' But then so was Μαγδῶλα, the name of a village in 
the Fayum. Yet the reader who ransacks the volume of papyri found 
at Magdola ? will find as many Hebraisms there as he would find Gallicisms 
in a bundle of letters written by a farmer of Ashby-de-la-Zouch. It is 
in vocabulary that the Egyptian papyri show any such influence. Even 
here, as Schubart remarks, but few foreign loan-words are to be found. 
There are several Semitic and Persian terms, even fewer Latin words 
than we have in the Gospels, in spite of the large number of Romans and 
Italians living in Egypt, and, most remarkable of all, the native Egyptian 
languages have contributed practically nothing.* 

The other form of the argument emphasises the ubiquity of the Semitic 
stock. 

“Some dialect of the Semitic family was spoken in nearly all the 
eastern regions subdued by Greeks and Romans. ... The mixing of 
languages in the different provinces at this time will tend to explain the 
peculiar terms and formulae found in Aramaic and Greek inscriptions 
and papyri dating from the Imperial age. Many of these linguistic 
idioms may still be properly called ὁ Semitisms,’ though no longer found 
exclusively in the LXX and the NT. We may reject the notion of a 
* Judaeo-Greek’ dialect, but the Greek received something of its pre- 
valent form ἡ by its passage through the Semitic mind.’ . . . The Greek 
papyri show how Semitic influence was perpetuated in Greek-speaking 
regions. Many ‘ Semitisms’ silted through into the Κοινή, and became 
naturalised in the Greek vernacular.” ὃ 

This appears to be an overstatement for which no adequate evidence 
has yet been adduced. If, however, it were to be fully substantiated, 
this would bea striking confirmation of Deissmann’s claim that the Greek 
of the New Testament is essentially the spoken Κοινή of the world of 


1 P Fay 108%. 2 Papyrus Grecs de Lille, tome ii. 

3 Hinfithrung in die Papyruskunde, 188 f. 

4Thackeray dealt with this question, Gr. 20. See also Mayser 
Gr. i. 35-43. Later writers confirm the statements of Thumb Hellen. 
107-120. 

5 J. Courtenay James, The Language of Palestine, 70 ft. Unfortunately 
this argument is confused by the introduction of NT Semitisms. On the 
main issue, however, Mr. James has not made his position clear. He appears 
to agree with Harnack and Moulton that some natural productions of the 
Kow7 more or less accidentally coincide with Semitic forms. Six examples 
are given. ‘ But even in these instances the prominence and frequency of 
such words and phrases must be attributed to a nearer or more remote 
Semitic influence.’ Five other constructions equally attested are then given 
as neither natural developments of the Greek, nor accidental coincidences 
with Hebrew or Aramaic. ‘‘These and many other forms found in Greek 
inscrip. and papyri could scarcely have come into the Kow7 except through 
Semitic.’ A complete list of such forms, with accurate information as to 
date and provenance, would provide a valuable test for Mr. James’s theory. 
Meanwhile, it is difficult to recognise any principle behind the classification 
observed in n.* and n.* on p. 72. 


416 A GRAMMAR OF NEW TESTAMENT GREEK. 


Hellenism. The only difference would be that whereas the one school 
finds this unity in the comparative absence of distinctive Semitisms, the 
new school would have to concede this identity by postulating Semitism 
as a widespread characteristic of popular Hellenistic. Fortunately we 
have not to choose between these alternative routes to the same goal. 
We have rather to inspect every trace of possible Semitic influence in the 
Greek of the New Testament in order to determine the degree of proba- 
bility that any book has reached us through a Semitic medium, whether 
near or more remote. 

The method of treatment will be as follows. As far as possible all 
passages will be tabulated which Semitic scholars have claimed as obvious 
or possible Semitisms. They will be classified under grammatical head- 
ings, the passages under each heading being arranged in four groups, 
distinguished by numbers in bold type thus: 1. the Synoptic Gospels and 
Acts, 2. the Johannine writings, 3. the Pauline Epistles, 4. the remaining 
Epistles. The distribution of these constructions throughout the New 
Testament can thus be seen at a glance. The distinction between 
Hebraism and Aramaism will be observed wherever it should be recorded, 
and separate notice will be taken of passages where a difficulty in the 
Greek might be resolved by reference to a hypothetic Semitic original. 
Evidence will be offered, where it is available, of parallel usage in literary 
Greek, or in the Κοινήῆ. To this extent the Appendix may serve as a 
fairly comprehensive and critical survey of the present stage of the dis- 


cussion. It is an attempt to assemble the data upon which students 
will form their own judgments. 


A. GENERAL STYLE AND STRUCTURE OF SENTENCE. 
1. ῬΟΒΙΤΙΟΝ OF THE VERB IN THE SENTENCE. 


An important consideration urged by Wellhausen (W1 18 f., 510 1.) 
is the Semitic order of words. He claims that in Mark, with few excep- 
tions, the verb is followed by the subject. This argument has been 
seriously discounted by Lagrange (8. Marc, Ixxxviii) on three grounds : 
(a) This order is Hebrew and Arabic rather than Aramaic and Assyrian 
(e.g. in Aram. portions of Daniel the verb more often follows than precedes 
the subject). (b) Whilst the verb does more often precede than follow 
the subject, if those instances are taken into account where, under the 
form of a participle, the subject really precedes, the balance is changed. 
A statistical examination of Mk 1-2 shows the subject after the verb 27 
times, the subject before the verb 18, sentence beginning with participle 
18. Moreover, in the Passion narrative the verb is far less often before 
thesubject. (c) In Greek the verb is often enough put before the subject, 
with the same rule as in Semitic, of leaving it in the sing. when it has 
two subjects. However vaguely Semitic the order of words may be 


im Mark, it does not in this respect give the un-Greek impression of 
1 Maccabees, 


SEMITISMS IN THE NEW TESTAMENT. 417 


Strangely enough, Wellhausen, as we have already observed (p. 32 
above), does not attribute the priority of the verb in the word order of the 
Fourth Gospel to Aramaism, but rather to general imitation of the 
Biblical style, an explanation, which, in Prof. Torrey’s opinion, “ has 
not the least plausibility” (H7'R xvi. 323). Burney is silent upon the 
question. 

Strong support comes from EK. Norden, for the judgment of such a 
Hellenist as the author of Die antike Kunstprosa carries great weight. 
“Placing the verb first is, next to parallelism of clauses—the two are 
very often combined—the surest Semitism of the NT, especially in those 
instances in which this position comes in a series of clauses. That has 
struck me specially in Luke among the Synoptists, although in other 
respects he inclines more to the Hellenic side. But he shows a greater 
preference than the others for a flowing style of narrative, in which there 
was naturally more opportunity for this position of the verb” (Agnostos 
Theos 365). Lk 15"#- is cited as illustrating the Semitic style of narrative, 
both by the position of the verb, and by the linking of parallel clauses 
with καί. The second half of the Magnificat is given as an instance’ 
of the priority of the verb without the repetition of καί. Of course the 
Hebraic style of these first two chapters of Luke is undeniable. The 
initial place of the verb in the series of clauses (bound by the common 
relative pronoun ὅς) in the credal hymn of | Ti 3!° is declared to correspond 
with its Hebraic parallelism of thought (7b. 257). -So also the distinctive 
position of the imperatives in the Lord’s Prayer is compared with the 
style of Jewish prayers, as in Is 3717-79, Sir 36117, 

But the subject calls for more exact treatment than this, and Thum)h’s 
section on ‘‘ Wortstellung,” appended to Brugmann’s Gr. Gr.* 658 f1., is a 
useful corrective. Thumb points out a distinction observed in MGr 
between the order of words in a principal sentence and in a subordinate 
clause, and thinks that in the Κοινή this distinction was beginning to 
assert itself. “‘ In dependent clauses without exception the verb follows 
immediately upon the introductory particle, or is separated from it only 
by the negative or the conjunctive pronoun, and practically without 
exception the verb follows upon an interrogative word and a relative ” 
(Handb. 202). We have therefore to give special attention to the place 
of the verb in principal clauses. Here he rejects the rule for classical 
Greek given in K-G ii. 595: ‘‘ the subject takes the first place, the pre- 
dicate the last, and the object comes before the predicate.” Delbriick, 
with more caution, says, “‘ We have the general impression that the posi- 
tion is a matter of freedom”? (Grd. v. (iii.) 65). Emphasis could, of course, 
be gained by putting any word out of its usual order. The fullest in- 
vestigation so far is that by Kieckers (see above, 32 n.), who examines the 
place of the verb in principal clauses (excluding direct imperatives) in a 
selection of passages from several classical and Hellenistic prose authors. 
For this purpose he has taken 20 pages from Herodotus, Thucydides, 
Polybius, and the Chronicle of Theophanes, 20 also from Xenophon (10 
from the Anabasis and 10 from the Hellenica). Five pages are also taken 
from each of the four Evangelists. The position of the verb is classified 


418 A GRAMMAR OF NEW TESTAMENT GREEK. 


according to its occurrence in the initial (1), middle (M), end (E) position. 
The results are : 


| I M E 
Eaves fe ao. | ἘΞ eee α ν, Ὁ 
Thadedidas 55. πο πη δὺο Ὁ ἀπ με ον 
Xenophon, Anab. . : : | 34 | 95 45 
Hell. : Ἢ 10 89 48 
50 —184. —93 
ice ee ier Fyn io ate gee ὦ 
ΠΣ ΙΣΣ ΝΣ 
Mark |. icq Reupietnlt lk ae Mineo με 4 
Lake? =? eae “Hegel Mine ABA sen 
Tee SSN PE a pe eae es ete | 
“REG phanes πὶ ον ΩΝ Ὡς 
{ 


This high ratio of verbs in the initial position in the Gospels is largely 
due to the considerable number of verbs of saying, which in accordance 
with regular Greek usage stand at the beginning of their sentence. 
Kieckers proves (p. 64) by numerous examples how very common this 
is in the simple narrative of Herodotus. It is very rare in the more 
dignified style of Thucydides, where important political speeches are 
generally introduced with more elaborate phraseology. Brugmann (KVG 
683) states that the initial position of the verb is usual throughout I.E. 
speech, at the beginning of a narrative, or in the carrying on of the 
narrative in a recital consisting of a series of clauses. This order in 
narrative is less common in Skt. and Gr. than in Germ. and Slav., but 
more common than in Lat. When all allowance has been made for 
these factors, the predominance of initial position in Luke and John is 
remarkable. 


2. PARALLELISM. 


The questions raised under this head can hardly be treated in a 
Grammar of NT Greek. A reference to the principal authorities must 
suffice. It has long been recognised that much of the teaching of Jesus 
falls into the poetical style of Hebrew poetry and Wisdom Literature. See 
C. A. Briggs’ “The Wisdom of Jesus the Messiah” (Exp T' viii. 393-398, 
492-496, ix. 69-75). Burney (J7'S xiv. 414 ff.) showed that the Parable 
of the Last Judgment if turned into Hebrew is a rhythmical structure 
which is largely lost when rendered into Aramaic. But in his post- 
humously published work, The Poetry of our Lord, an elaborate proof is 


SEMITISMS IN THE NEW TESTAMENT. 419 


attempted that very much of the teaching, when translated into Aramaic, 
conforms to the various metres of the poetry of the OT. It is for Semitic 
scholars to say how far a valid argument can be built upon such data for 
the contention that Mt preserves the teaching of Jesus as given in Q 
better than Lk. Inasmuch as the teaching of Jesus, and the original 
record of it, is assumed throughout this Grammar to have been in Aramaic, 
and since the Semitic colouring of our Greek documents in these very 
parts is not in dispute, no more need be said. For the influence of the 
LXX upon Paul’s most exalted passages, and for a study of parallelism 
in the Epistles, see J. Weiss Festschrift f. B. Weiss, P. Wendland Die 
urchristlichen Literaturformen 355f., and the appendix on ‘ Semitischer 
und hellenischer Satzparallelismus,”’ in E. Norden Agnostos Theos 355 ff. 


3. TAUTOLOGY. 


A Semitic colouring is seen in the continual repetition of an idea by 
(a) a subordinate clause, or (Ὁ) a co-ordinate parallel sentence. Thus : 
(a) Mk 1713 (cf. Mt 156), Mk 128 (cf. Mt 2228), Mk 1319 (cf. Mt 241). 
(Ὁ) Mk 2195 (cf. Mt 915), Mk 430 (Ὁ ἐν ποίᾳ παραβολῇ παραβάλωμεν) 
(ΜΠ 1931. Ek 1315}. Mike 1135 ((=Mt.). 213 Shik 202); 
Mk 1214 (cf. Mt 221”, Lk 2035). 

Archdeacon Allen, who gives a much fuller list (Hxp 7 xiii. 329), shows 
that in all these examples Mt has pruned some of the redundancy. The 
faithful rendering of the Aramaic teaching of Jesus is responsible for 
many “ symmetrical tautologies’’ in non-Marcan sources (W! 18), e.g. 
MiG Mt 6 (— kk 1G) Mb 15: (— Ek 64) Mb 74a (— Tk ΤΕ Mit 
713% Mt 1115 (toned down in Lk 16!*). There is, of course, no violence 
to Greek idiom in these instances. 

A special form of tautology is that accompanying the adverbial use 
of the relative pronoun (which is often attracted into another case) : 


1. Mt 279 τὴν τιμὴν τοῦ τετιμημένου ὃν ἐτιμήσαντο ἀπὸ υἱῶν Ἰσραήλ. 
Not from LXX of Zech 1115; probably from Testimonies. 
See J. R. Harris, Testimonies, i. 58 f. Also McNeile, Lagrange, 
Allen in loc. 
2. Jn 1755 ἡ ἀγάπη ἣν ἠγάπησάς pe. 
3. 1 ΤῊ 3° ἐπὶ πάσῃ τῇ χαρᾷ ἧ χαίρομεν. 
1 Co 739 ἕκαστος ἐν τῇ κλήσει ἧ ἐκλήθη ἐν ταύτῃ μενέτω. 
Eph 15 χάριτος ἧς ἐχαρίτωσεν, 115 ἐνέργειαν ἣν ἐνήργηκεν (cf. 339, Col 
159), 24 ἀγάπην ἣν ἠγάπησεν, 41 κλήσεως ἧς ἐκλήθητε. 
(Eph 3:9 44, 2 Ti 1° illustrate a different tautology.) 


Dr. J. Rendel Harris, who called attention to this characteristic of 
Eph. (letter to J. H. M., Dec. 28, 1913), also suggests that some of these 
‘“* Aramaisms”’ have been corrected, and that 18 τὸ πλοῦτος τῆς χάριτος 
ἧς ἐπερίσσευσεν, had 6 for ἧς, 15 εὐδοκίαν ἣν προέθετο was originally 
πρόθεσιν ἣν προέθετο, 34 κατὰ πρόθεσιν... ἣν ἐποίησεν was ἣν προέθετο 
(the pleonasm having been resolved in two different ways). 


420 A GRAMMAR OF NEW TESTAMENT GREEK. 


He further urges large elements of Aramaism in 2 Peter. 


4. 2 Pet 38 ἐμπαιγμονῇ ἐμπαῖκται, 212 ἀδικούμενοι μισθὸν ἀδικίας, 21% 
ἡδόνην ἡγούμενοι τὴν ἐν ἡμέρᾳ τρυφήν (regarding the almost 
meaningless ἡγούμενοι as a substitute for the pleonastic 
ἡδόμενοι). “ἡ The whole sentence is pleonastic and Aramaic.” ἢ 
P. Wendland also (Die urchristl. Literaturformen, 369 n.?) 
observes that 2 Pet 215 3° give an impression of Semitism not 
found in the parallels in Jude. 


In view, however, of Epict. i. 29. 49, ταῦτα μέλλεις μαρτυρεῖν καὶ 
καταισχύνειν τὴν κλῆσιν ἣν κέκληκεν (ὁ Geos) . . . 3 it seems needless to 
label this idiom Semitic. 


4. PARATAXIS. 
Under this heading we must bring :— 


(a) Co-ordination of clauses with the simple καί, instead of the 
use of participles or subordinate clauses.—This is far more common in 
Mk than in either Mt or Lk. As a characteristic of John, see Burney 
Aram. Orig. 56. 

Wellhausen (W!? 21, 225), who regards B as more reliable than D for 
particles, shows that δέ has often been substituted for Mk’s καί by 
the other evangelists or by copyists in all Gospels. He further claims 
that there is evidence in the MSS of resolved, or imperfectly resolved, 
parataxis in a number of passages, e.g. D retains parataxis Mk 3”! (ζκουσαν 
καὶ ἐξῆλθον), similarly 435 810 1055, Also Lk 223% σὺ δὲ ἐπίστρεψον καὶ 
στήρισον. In Mk 67:15 B retains parataxis where D has participle. 
D sometimes links a participle and a finite verb with καί. H.g. Mt 2613 
πορευθεὶς. . . καὶ εἶπεν, Lk 9° ἐξερχόμενοι... καὶ ἤρχοντο. There are 
many such instances in the D text of Mark, but, as Lagrange has shown 
(S. Marc, p. lix), these are cases (Mk 557 725 1]? 141-63) where καί has 
been inserted to correspond with the number of Latin words in ὦ (1614 d 
is missing). 

For bearing of this upon general question, see Proleg. 12. 

Milligan (Vocab. s.v. καί) thinks it “‘ impossible to deny that the use 
of καί in the LXX for the Heb. } influenced the Johannine usage.” 
Lagrange, in view of the slight trace of LXX influence on Jn, suggests 
Aramaic for Heb. For the hypotactic force of καί from Aristotle to 
MGr, see Thumb Hellen. 129, and for examples of some of the following 
usages, Thumb Handb. 184. 


(Ὁ) The co-ordinate use of subjunctive after (a) an imperative, 
or (8) θέλειν. 


E.g. (a) Mk 1** ὅρα μηδενὶ εἴπῃς (but see Vocab. 455 δ). 
Mt 7! ἄφες ἐκβάλω (but see Prol. 175. Common in Epict., e.g. 
1. 9. 15, ἄφες δείξωμεν αὐτοῖς ; ii. 18. 24, ἄφες ἴδω τίς εἶ). 


1 In the absence of any textual warrant one must resist the tempting 
suggestion to read 2 Pet 213 ἐν φθορᾷ φθαρήσονται. 


SEMITISMS IN THE NEW TESTAMENT. 421] 


(38) Mk 1039 τί θέλετέ με ποιήσω ὑμῖν (SB, numerous v.ll.), Mk 105 
μ 
(=Mt 20°=Lk 183), Mk 141? (=Mt 26"7--ΞΔῇ Κὶ 229), Mk 15:2 
(ADe), Lk 954, Mt 1328. 


Plummer [CC Luke, p. 264, who notes that ἵνα is not inserted 
when the first verb is in the second person, and the second verb in the first 
person, cft. Soph. OT 650. In class. Gr. common with βούλομαι, which 
is largely replaced by θέλω in NT. Cf. Plato Gorg. 521 d, βούλει σοι 
εἴπω; Xen. Memor. τι. i. 1, βούλει σκοπῶμεν ; 10, βούλει σκεψώμεθα; 
cf. Lat. vis maneamus? See Prol. 185. Cf. BGU i. 38 (ii/ili Α.}.) τί 
θέλεις ἀπενέγκω αὐτῷ. 

There is therefore nothing Semitic in Jn 1839 βούλεσθε οὖν ἀπολύσω 
ὑμῖν τ. Bao. τ. Ιουδαίων ; 

(6) The conditional parataxis of the imperative——Wellhausen dis- 
covers a Semitic locution when two imperatives linked by καί represent 
the protasis and apodosis of an implied condition (W1 25). 


1. H.g. Mk 834 (=Mt 16%4*=Lk 9528) ἀπαρνησάσθω ἑαυτὸν καὶ ἀράτω τὸν 
σταυρὸν αὐτοῦ καὶ ἀκολουθείτω μοι (=then he will be my disciple), 
Lk 7? εἰπὲ λόγῳ, καὶ ἰαθήτω ὁ παῖς μου (=Mt 88... καὶ 
ἰαθήσεται). Α 1685 striking example is the imperatival protasis 
followed by καί and the future; eg. Mt 7’ (=Lk 11%), Lk 
1058 τοῦτο ποίει καὶ ζήση. Wellhausen adds Mt 1238 with the 
remark, “‘ The καί introduces the apodosis, and after this the 
nominative would have been used more fittingly—xai ὁ 
καρπὸς αὐτοῦ καλός. He recognises, however (W? 13), that 
this idiom, ‘‘ Divide et impera,” “᾿ Give a dog a bad name 
and hang him,” is found in all languages. It is certainly: 
good Greek, as in Soph. H/. 1207, πιθοῦ λέγοντι, κοὐχ ἁμαρτήσει 
ποτέ. For other exx. see K-G ii. 248. [MGr.—R. McK.] 


Burney describes this as less characteristic of Aram. than of Hebr., 
“ἢ except where the sequence is clearly to be regarded as the result of the 
preceding imperative.” For examples in Hebr. and Aram., see Aram. 
Orig. 95. He cites : 


2. Jn 189 ἔρχεσθε καὶ ὄψεσθε. 1624 αἰτεῖτε καὶ λήμψεσθες Elsewhere we 
PX Ἱ 
fin 


Rev 4! avaBa ὧδε, καὶ δείξω σοι. 
4, Jas 45. & 10, 


(4) The temporal use of καί in parataxis—(W!. 20. In 713 re- 
cognised as quite good Greek.) 


1. H.g. Mk 15” ἣν δὲ ὥρα τρίτη καὶ ἐσταύρωσαν αὐτόν. 
IL 9344 Ne HS ς \ ef “ \ , Sees 
4k 295 καὶ nv ἤδη ὡσεὶ ὥρα ἕκτη Kal σκότος ἐγένετο. 
43 Ὁ a c , > Ν ‘ ‘ , ,ὔ 
19% ὅτι ἥξουσιν ἡμέραι ἐπὶ σὲ καὶ περικυκλώσουσίν σε. 
Mt 26% ἰδοὺ ἤγγικεν ἡ ὥρα καὶ 6 vids τοῦ ἀνθρώπου παραδίδοται. 


τ We might add “ Wait and see” (-- ἐὰν μείνητε, ὄψεσθε) in its historic 
use in the British House of Commons. 


422 A GRAMMAR OF NEW TESTAMENT GREEK. 


On this see Proleg. 12 n.2.. Thumb, in Brugmann Gr.‘ 640, eft. Xen. 
Anab. τι. 1. 7: καὶ ἤδη τε ἦν περὶ πλήθουσαν ἀγορὰν Kai ἔρχονται . . 
κήρυκες, and deprecates the suggestion of Hebraism. Bauer (Lex. 611) 
gives several reff. to classical authors. [Common in MGr.—R. McK.] 

We may add an interesting example from Tob 18X (on which see 
D. C. Simpson OA i. 181, 203): ὅτε ὀρφανὸν κατέλιπέν pe ὁ πατὴρ καὶ 
ἀπέθανεν (“ἡ Because my father left me an orphan when he died”’). 

Other possible instances in the NT are: 


Did eae 22° ASS γνο- 
4, Heb 88 (LXX). 

(e) The consecutive use of καί in parataxis (closely related to (c) 
and (f). (See Lagrange S. Matthieu p. xc f., S. Jean p. evii, Burney Aram. 
Orig. 68. For Waw apodosis, see Ges-K, ὃ 143 (d), Kautzsch Aram. Gr. 
8 69. 1, Marti K.Gr. 105 1.) 


1. Mt 64 καὶ ὁ πατήρ σου... ἀποδώσει σοι. 
82: ἐπίτρεψόν μοι ἀπελθεῖν καὶ θάψαι (contr. Lk 98 ἀπελθόντι 
θάψαι). 
Lk 25: καὶ ὅτε ἐπλήσθησαν ἡμέραι... καὶ ἐκλήθη τὸ ὄνομα αὐτοῦ Ἰησοῦς. 


2. Jn 510 67 1148 1416. Lagrange cft. Plato Phaed. ὅθε, καὶ ἥκομεν καὶ 
ἡμῖν ἐξελθὼν ὁ θύρωρος εἶπε. 
The clearest instances are in the Apocalypse. See Charles 7166 
i. 101, 265, 11. 16. 


Rev 950 x 046. ἐάν τις ἀκούσῃ τῆς φωνῆς μου... καὶ ἐλεύσομαι πρὸς 
αὐτὸν. ᾿ς: 
10? ὅταν μέλλῃ σαλπίζειν, καὶ ἐτελέσθη τὸ μυστήριον τοῦ θεοῦ... 
149% 10 εἴ τις προσκυνεῖ τὸ Onpiov . .. καὶ αὐτὸς πίεται ἐκ τοῦ 
olvov. ... 
3. Phil 1°? εἰ δὲ τὸ ζῆν ἐν σαρκὶ τοῦτό μοι καρπὸς ἔργου, καὶ Ti αἱρήσομαι ; 
οὐ γνωρίζω. (But if... , then what shall I choose 3) 


Radermacher (Gr.? 223) accepts this reading and punctuation, but counts 
it vernacular rather than translation Greek. [MGr.—R. McK.] 


(f) Interrogative parataxis, where καί introduces (a) a temporal 
apodosis (W1 20, 713), closely related to (e); or (8) a paradox (Burney 
Aram. Orig. 67). 

(a) 

1. E.g. Mt 187! ποσάκις duaprnoe . . . ὁ ἀδελφός μου καὶ ἀφήσω αὐτῷ; 
2055, Ος͵ ὅτι οὐ δύναμαι παρακαλέσαι τὸν πατέρα μου καὶ 
παραστήσει μοι... -; 
Lk 145 τίνος ὑμῶν υἱὸς ἢ βοῦς εἰς φρέαρ πεσεῖται καὶ οὐκ εὐθέως 
ἀνασπάσει αὐτόν. . .; 
2436 οὐχὶ ταῦτα ἔδει παθεῖν τ. χριστὸν καὶ εἰσελθεῖν. . .; 
3. Rom 1135 (LXX). 

(3) 

2. Jn 279 τεσσεράκοντα καὶ ἕξ ἔτεσιν οἰκοδομήθη ὁ ναὸς οὗτος, καὶ 
σὺ ἐν τρισὶν ἡμέραις ἐγερεῖς αὐτόν; So 310 867 934 118, 


SEMITISMS IN THE NEW TESTAMENT. 423 


Lagrange adds 1251 ἡμεῖς ἠκούσαμεν ἐκ τοῦ νόμου... Kal πῶς λέγεις 
σύν wee swith the comment that this interrogative phrase beginning 
with καί, “‘ and yet” (often ironical), was quite good Greek. Cf. Eur. 
Medea 1398, κἄπειτ᾽ ἔκτας ; [MGr.—R. McK.] 

For καί adversative see below, under Conjunctions (p. 469). 


(g) Cireumstantial clauses introduced by καί. (See W! 19, Ev. 
Marci 36, Hv. Lucae 110. Ges-K 8 156). [Common in MGr.—R. McK.] 


1. Mk 1'9 καὶ προβὰς ὀλίγον εἶδεν Ἰάκωβον... καὶ Ἰωάννην τ. ἀδελφὸν 
αὐτοῦ, καὶ αὐτοὺς ἐν τῷ πλοίῳ καταρτίζοντας τὰ δίκτυα. 
431 καὶ καθεύδῃ καὶ ἐγείρηται νύκτα καὶ ἡμέραν, καὶ ὁ σπόρος 
βλαστᾷ καὶ μηκύνηται. 

Lk 195 καὶ ἐδαφιοῦσίν σε καὶ τὰ τέκνα σου ἐν σοί (1.6. the enemy 
will beleaguer the city while her children are in her— 
not only the regular population, but all Jews from far 
and near who have fled to the metropolis). But Torrey 
(C. H. Toy Studies, p. 283 n.) shows from Nah 319, Hos 1014 
141, that these words are the object of the verb, not a 
circumstantial clause. 


Charles (JCC i. p. exlviii; ii. 120, 417, 431) thus explains καί (= 
seeing that) in three passages in the Apocalypse : 
2. Rev 12" καὶ αὐτοὶ ἐνίκησαν αὐτὸν διὰ τὸ αἷμα τοῦ dpviov. 
18% καὶ οἱ βασιλεῖς τῆς γῆς μετ᾽ αὐτῆς ἐπόρνευσαν. 
93 ᾿Αλληλούϊα' καὶ 6 καπνὸς αὐτῆς ἀναβαίνει εἰς τοὺς αἰῶνας τῶν 
αἰώνων. 

(h) On καὶ εὐθύς in Mark (=otv in John)=Waw Consecutive, see 
Burkitt Hv. da-Mepharreshe, ii. 89, Burney Aram. Orig. 68 τ. 

On parataxis as a test of Semitism, see also E. Norden Agnostos Theos 
367, with Deissmann’s reply LAH? 132 ἢ. Radermacher (Gr.? 218) 
cites many parallels from later Greek writers, and concludes that this 
was a feature common to the popular speech in Hebrew and Greek. 


5. Casus PENDENS, FOLLOWED BY RESUMPTIVE PRONOUN. 


This is not to be confused with the construction in which the subject 
of an interrogative sentence is put first for emphasis or clearness, 6.0. 
Mk 1139 τὸ βάπτισμα τὸ Ἰωάνου ἐξ οὐρανοῦ ἢν ἢ ἐξ ἀνθρώπων ; 

The casus pendens is generally, but not always, the nominative ; the 
resumptive pronoun may be in any case. This usage is specially fre- 
quent with participial clauses. Here, again, D seems to preserve the 
construction in several passages where most MSS have smoothed away 
the irregularity. 

1. Two OT citations can be illustrated from the Hebrew original : 
Mk 12!°(=Mt 212=Lk 2017) λίθον ὃν ἀπεδοκίμασαν οἱ οἰκοδομοῦντες, οὗτος 

ἐγενήθη εἰς κεφαλὴν γωνίας. (Cited from Ps 117(118)??.) 
138 NAD nny Ovi AND Ἰ3 


Here the Greek syntax is sis peated by the Hebrew. It is 


424 A GRAMMAR OF NEW TESTAMENT GREEK. 


otherwise in Mt 416 ὁ ads ὁ καθήμενος ἐν σκότει φῶς cidev μέγα, Kai τοῖς 
καθημένοις ἐν χώρᾳ καὶ σκιᾷ θανάτου φῶς ἀνέτειλεν αὐτοῖς. (ΟἿ, Is 93.) 
SOMSy maz TiN ΠΉΘΟΝ PID 3) Oa ΝΣ ae] ABI ὈΒΟΤΠ. oN 
For Casus pendens in Synoptic Gospels see W! 19 f., 711 f. 
Other examples are: 
Mk 134 D καὶ τοὺς δαιμόνια ἔχοντας ἐξέβαλεν αὐτὰ an’ αὐτῶν (where 
Wellhausen suspects an original οἱ δαιμόνια ἔχοντες). 
Mk 616 720 134, 
Mt 5% καὶ τῷ θέλοντί σοι κριθῆναι καὶ τὸν χιτῶνά σου λαβεῖν (Ὁ ὁ θέλων τὸν 
χιτῶνά σου λαβεῖν), ἄφες αὐτῷ καὶ τὸ ἱμάτιον. 
101: D ἡ πόλις, εἰς ἣν εἰσέλθητε εἰς αὐτήν, ἐξετάσατε τίς ἐν αὐτῇ. . .. 
¥12°° πᾶν ῥῆμα. . . ἀποδώσουσιν περὶ αὐτοῦ λόγον. 
1713 Ὁ καὶ ἐλθὼν πρὸς τὸν ὄχλον προσῆλθεν αὐτῷ ἄνθρωπος γονυπετῶν... .. 
1320. 22. 23 (- Tk 814. 16) 1338 1511 1928 9418 9529 9623, 
*Lk 12% παντὶ δὲ ᾧ ἐδόθη πολύ, πολὺ ζητηθήσεται παρ᾽ αὐτοῦ, καὶ ᾧ 
παρέθεντο πολύ, περισσότερον αἰτήσουσιν αὐτόν. 
Here the casus pendens is in the dative by inverse attraction of the 
relative. 
215 ταῦτα ἃ θεωρεῖτε, ἐλεύσονται ἡμέραι ἐν ais οὐκ ἀφεθήσεται λίθος ἐπὶ 
λίθῳ ὃς οὐ καταλυθήσεται. 
Cf. the parallels Mk 135, Mt 3945, in which the casus pendens has been 
absorbed by making it the object of βλέπειν. 
Other examples are Lk 121°* 134 235952, Acts 74° (LXX), 


* N.B.—Passages in which the suspended clause contains πᾶς. 


2. For casus pendens in Johannine writings, see Burney Aram. Orig. 
34, 63 ff., 151; Torrey H7'R xvi. 322 f.; Lagrange S. Jean cxf.; Charles 
ICC i. pp. exlix, 53. 

Jn 1 ὅσοι δὲ ἔλαβον αὐτόν, ἔδωκεν αὐτοῖς. . .. 

Burney cites also 118. 33 326. 32 51. 19, 36. 37, 38 639- 46 718 825 10!- 25 
1245: τὸ ais 18. ata 1 54-78-21 Ste ἢ “these, "Oo aloe ie awe ones 
in the suspended clause. 

1 Jn 24 ὑμεῖς ὃ ἠκούσατε ἀπ᾽ ἀρχῆς, ἐν ὑμῖν μενέτω. 
Rey 276 912. 2! ὁ νικῶν δώσω αὐτῷ. 68 ὁ καθήμενος ἐπάνω αὐτοῦ, ὄνομα 
αὐτῷ ὁ Θάνατος. In oblique cases 27- 17 64 218 (046). 

Lagrange shows that in several of the examples from Jn there is a 
degree of emphasis which accords with classical usage, and he regards 
them not as translation Greek but as locutions which would come naturally 
to those accustomed to the vigorous rather than varied Semitic idiom. 
But he recognises a Semitic locution when a pronoun resumes the clause 
introduced by πᾶς (δ΄. Jean, p. exi). 

Closely allied to this construction is one in which πᾶς is used with a 
participle in an introductory circumstantial clause, though no resumptive 
pronoun follows (W? 11). Such hyperbaton, however, is not unclassical. 
E.g. Lk 6* πᾶς ὁ ἐρχύμενος πρός με. . . ὑποδείξω ὑμῖν τίνι ἐστὶν ὅμοιος. 

(In Mt 7*4 the πᾶς ὁ ἐρχόμενος becomes subject οὗ ὁμοιω- 
θήσεται, and the syntax is more precise. Wellhausen 


SEMITISMS IN THE NEW TESTAMENT. 425 


suspects a similar correction of an original nominativus 
pendens in the gen. abs. of Mt 1319). 
The Hebrew construction (see Ges-K § 116 ~) may be illustrated by 
1 Sam 213 59 5) Nay Mat nat ΧΕΙ which, rendered quite literally 
into Greek, would be πᾶς θύων θυσίαν, ἤρχετο ὁ παῖς Tov ἱερέως. (The 
LXX, with a different verse division, reads καὶ τὸ δικαίωμα τοῦ ἱερέως 
παρὰ τοῦ λαοῦ παντὸς τοῦ θύοντος, Kal ἤρχετο TO παιδάριον τοῦ ἱερέως . . .). 
In 1 Sam 31} the same construction is rendered by gen. abs. in LXX. 
Closely akin to this is 


Sir 31°! θυσιάζων ἐξ ἀδίκου, προσφορὰ μεμωκημένη, Kat οὐκ εἰς εὐδοκίαν 
μωκήματα ἀνόμων. 
(RV 34:8 He that sacrificeth of a thing wrongfully gotten, his offering is 
made in mockery, 
And the mockeries of wicked men are not well-pleasing.) 


Box and Oesterley (0.4. i. 435) supply an original nmap Syn nat 
ὌΝ, pointing Say nat, where the Greek translator read rn nat, 
and rendering “The sacrifice of the unrighteous man is a mocking 
offering.” 

The casus pendens, followed by resumptive pronoun, is said by 
Radermacher (Gr.? 21f.) to appear quite early in Greek literature and 
to be common in the later language. His examples from Hom. Od. 
xii. 73 and Xen. Cyrop. I. vi. 18 are not parallel, but the Silco inscr. 
(OGIS 2011%) is near enough: οἱ δεσπόται τῶν ἄλλων ἐθνῶν οὐκ 
ἀφῶ αὐτοὺς καθεσθῆναι εἰς τὴν σκιάν. Quite a crop of instances has been 
gathered from Aelian De Nat. Anim. e.g. 1. 48, ὁ κόραξ, ὄρνιν αὐτόν. 
φασιν ἱερόν. Cf. 1. 19. 55, ii. 51. 

For other instances see K-G i. 47. 660. (The two cited by Mr. 
G. R. Driver, Lucian, Dial. Mort. xii. 5, Epict. Ench. 42, are not parallel.) 
ἐνεγκών) σοι THY ἐπιστολήν, δὸς αὐτῷ ἄλλην. See also Thumb Hellen. 131, 
and, for survival in Μόν, Handb. 32. [Dr. McKinlay shows that the 
idiom is so thoroughly vernacular that, out of 27 instances cited by 
Burney in Jn, Pallis retains 25 in his Romaic Gospels. There is no 
parallel in MGr to the nom. pend. with part., but the constr. was 
common in Med. Gr. | 


6. CONSTRUCTIONS WITH καὶ ἐγένετο. 


These are (a) καὶ ἐγένετο ἦλθε, (Ὁ) καὶ ἐγένετο καὶ ἦλθε, (0) ἐγένετο 
ἐλθεῖν. See Proleg.* 15 ff. for detailed treatment. 

Thackeray (Gr. 50 ff.) shows that with a single exception LXX uses 
(a) and (Ὁ) only. His statistical table proves that (6) predominates in 
LXX as does its equivalent in Hebr., but this predominance is accounted 
for by the slavish imitation of the Hebr. in the later historical books. 
The first two books of the Pentateuch and the prophetical books prefer 
(a). Both (a) and (6) seem to have been “ἡ experiments of the translators, 
which must he classed as Hebraisms,” but the asyndetic form (a) “° was 


426 A GRAMMAR OF NEW TESTAMENT GREEK. 


rather more in the spirit of the later language, which preferred to say, 
e.g., ᾿ Ft happened last week I was on a journey,’ rather than ‘It was a 
week ago and I was journeying.’’’ We notice further that while the 
translator of 1 Mac uses (a) three times and (6) five times, the free Greek 
of 2-4 Mac avoids both and retains the classical συνέβη c. infin. (as 
in Gen 41:3. 4238), 

The following tables (based partly upon the references supplied in 
Hawkins HS 37) will illustrate what has been written in Proleg. about 
the distribution of these constructions in the Gospels and Acts. 


(a) καὶ ἐγένετο (féyéveto δὲ) ἦλθε. 


Temporal Clause. 


NE ἡ cence eeseeetinasenetionnsentennmmimentantmie st 
| 
ἘΞ fe ὡς 6. aor, ὅτε c. aor. Other Time 
Pee mt: ind. | ind. Determination. 
Mark Ἢ 44 od | τ | 19 
ρῶν, | oS ie! : ἤπαν τοι Wai Je 
Matthew we si 72 111 7585 
191 261 
κ᾿ . 15} 351 918:88. 138. “1 91 er 19. 58 a6 αι] 
] GS a 1171: 1039 ΠΡ οι δ Ἢ is, | 
18394 2430-51 | 9091 
| | | 
—_ = ~ | [π ——— ) 
lets hs) τ ie | bss | δὰ [45 1: 


(0) καὶ ἐγένετο (Τ ἐγένετο δὲ) καὶ ἦλθε. 


Temporal Clause. 


EEE 


Sutin ink ws c. aor. | ὅτε 6: aor. Other Time 
: ind. | ind. Determination. | 

Mark 
Matthew . a ay = gio 
uke eae 1078} oe ΣῊ διῖ 81 Beep 

1111 [018 Ϊ 

)43. 16 | 

| La 

Ἀορ . - [22D] Se a ΕΠ" 


1 See Proleg.* 16 n.? 70. 233. 


SEMITISMS IN THE NEW TESTAMENT. 427 


(c) ἐγένετο δὲ (*xat ἐγένετο) ἐλθεῖν. 


Temporal Clause. 
ee Ὁ 


ἐν τῷ ο. inf. ὡς ο. ind. ὅτε c. ind. momen ne 

‘Mark .. a | ae ie 228% [215 ηίνεται 
| without temp. cl. ]. 

| Matthew 

|Luke . . 3% Be οἷς 811. (gen. abs.) 

] 61: 6 12 1622 (with- 

out temp. cl.) 
Acts . .| 93192 a . 45 982 87-43 1126 141 


| 1616 [211 5] 298: 17 
| | 274* (καὶ οὕτως ἐγ. 
without temp. cl.) 
| 288 (without 
| temp. cl.) 17, 


These constructions are thus distinctively Lucan, with a marked 
contrast between the Third Gospel and Acts, which becomes still more 
striking when we observe that in the latter Luke not only uses (c) almost 
entirely to the exclusion of (a) and (6), but also avoids the more Hebraic 
form of the time clause. Twice the familiar ἐν τῷ c. inf. occurs (once 
in the Palestinian narrative, once in the later story), four times a pre- 
positional phrase indicates the time, four times a participial construc- 
tion is used, twice an accus. of time duration modifies the infin., and 
twice there is no temp. clause at all. In other ways the construction 
tends to depart from the Hebraic pattern. In 98 the formula is changed 
to ἐν δὲ τῷ πορεύεσθαι ἐγένετο : it is hardly recognisable in 211 ὡς δὲ 
ἐγένετο ἀναχθῆναι ἡμᾶς .. . ἤλθομεν, or in 215 ὅτε δὲ ἐγένετο ἐξαρτίσαι 
ἡμᾶς τὰς ἡμέρας .. ., Still less in 1055 ὡς δὲ ἐγένετο τοῦ εἰσελθεῖν τὸν 
Πέτρον (with which Plummer JCC Luke, p. 45, after J. R. Lumby, cft. 
Acta Barn. 7, ὡς δὲ ἐγένετο τοῦ τελέσαι αὐτοὺς διδάσκοντας). 

The classical word συμβαίνω is used for γίνομαι in this sense once 
in Acts, possibly because γίνομαι has already appeared in the sentence : 
Ac 21% ὅτε δὲ ἐγένετο ἐπὶ τοὺς ἀναβαθμούς, συνέβη βαστάζεσθαι αὐτὸν 
ὑπὸ τῶν στρατιωτῶν. This is good vernacular Greek also, as we see 
from Tob 87 δὲ ἐν τῇ ἡμέρᾳ ταύτῃ συνέβη Zappa... καὶ αὐτὴν ἀκοῦσαι 
ὀνειδισμούς... . The equivalence of the two verbs in this sense in the 
Kown can be illustrated from papyri. Thus P Par 4959 (ii/B.c.) παρα- 
καλέσας αὐτὸν ἀπόστειλον πρὸς ἐμέ, γίνεται yap ἐντραπῆναι, where 
Witkowski (Hp. Ῥνῖυ.3 71) remarks, “ γίνεται c. inf.=ocvpBaiver c. inf. 

VOL. II. PART III.—29 


428 A GRAMMAR OF NEW TESTAMENT GREEK. 


Cf. Atticum ἔστιν c. inf. ut Plat. Rep. i. 3310. . . et ἔστιν ὥστε 
‘fieri potest, ut ; fortasse’ Sophocl.’’ See Vocab. 126a for pap. exx. of 
ἐὰν γένηται c. inf.=‘if it should happen that’ and σοὶ γίνοιτο c. inf. 
With these may be compared P Petr ii. 13(19)!° (iii/B.c.) as corrected 
by Wilamowitz (=Witkowski? p. 19) a[\V] ἔσται καὶ Θεύδωρον 
καταλειφθέντα ταὐτὸ ποιεῖν. Thackeray also calls attention to γίνεται 
εὑρεῖν = ‘it is possible to find’ in Theognis, and Xenophon’s use of ἐγένετο 
ὥστε or as = it happened that.’ 

Dr. G. G. Findlay (letter to J. H. M., December ’09) remarks: “The 
instances of ἐγένετο (δὲ) with infin. in the non-Palestinian parts of Ac 
seem worth considering separately, as evidence that the usage was not 
mere Hebraism. It is curious there is no non-Biblical ex. Ac 201 
seems decisive evidence of the native (or thoroughly naturalised) stamp 
of the idiom.” 


[(a) Common in MGr (with συνέβη, συνέβηκε), see Pallis’s Romaic 
Gospels. 


(b) Found occasionally in MGr dialects.—R. McK. | 


7. CO-ORDINATION OF THE PARTICIPLE OR INFINITIVE WITH THE 
FINITE VERB. 


(In addition to authorities cited below, see full treatment by Burney 
JTS xxii. 371-6.) 

Driver Hebrew Tenses ὃ 117, “* It is a common custom with Hebrew 
writers, after employing a participle or infinitive, to change the construc- 
tion, and, if they wish to subjoin other verbs which logically should be 
in the partep. or infin. as well, to pass to the use of the finite verb. Thus 
Gen 2738 δὲ 39) WS TWiT, ὁ θηρεύσας θήραν καὶ εἰσενέγκας (lit. ὁ Onpevoas 


θήραν καὶ εἰσήνεγκε). + The bearing of this upon the grammar of 
the Apocalypse was first shown by Archdeacon Charles (Studies in 
Apoc. 89 ff., ICC Revelation i. pp. exliv ff.), but Burney (Aram. Orig. 96) 
extends the usage to cover two examples in the Fourth Gospel, 
and quotes Dn 4%? to show that the construction is found in Aramaic 
also. 

On the other hand Holden, in his note on Xen. Cyrop. τι. iii. 8, 
cites 9 passages from the Cyrop. in which there is a transition from the 
participial construction to that of the verbum finitum. (The other 
eight are 1. ii. 5, τι. iii. 17, 21, m1. ni. 9, Iv. 11. 10, Vv. iii. 30, iv. 29, ΜΕ 
ii. 24). Cf. also Shilleto on Thue. τ. 57, 58, ‘“* The return from the sub- 
ordinate to the primary construction in Greek is too well known to require 
more than a passing illustration... . Iv. 100, ἄλλῳ τε τρόπῳ πειρά- 
σαντες καὶ μηχανὴν προσήγαγον (inst. of προσαγαγόντες). Plat. Theaet. 144c, 
ἀνδρὸς . . . καὶ ἄλλως εὐδοκίμου καὶ . . . κατέλιπεν (inst. of καταλιπόντος 
or ὅτι κατέλιπεν). Examples of this sort might be multiplied to any 


1 Dr. R. H. Charles JCC Rev. i. p. exlv, wrongly reters to this for a literal 
translation in LX.X. Even the reading of A εἰσήνεγκας does not secure that. 


SEMITISMS IN THE NEW TESTAMENT. 429 


amount.” For numerous examples see K-G ii. 100.‘ For later Greek 
see Jannaris Gr. ὃ 2168b. 

N.B.—In all these citations the participle is used for the verbum 
finitum, and in the great majority it is in the nominative. 


The NT occurrences are : 
(a) Participle. 
1. None. 
2. Jn 133 τεθέαμαι τὸ πνεῦμα καταβαῖνον. . . καὶ ἔμεινεν ἐπ᾽ αὐτόν. 
544 πῶς δύνασθε ὑμεῖς πιστεῦσαι, δόξαν παρ᾽ ἀλλήλων λαμβάνοντες, 
καὶ τὴν δόξαν τὴν παρὰ τοῦ μόνου θεοῦ οὐ ζητεῖτε (NW ζητοῦντες); 
2 Jn? διὰ τὴν ἀλήθειαν τὴν μένουσαν ἐν ἡμῖν, καὶ μεθ᾽ ἡμῶν ἔσται. 
Rev 15.5, τῷ ἀγαπῶντι ἡμᾶς καὶ λύσαντι ἡμᾶς... καὶ ἐποίησεν ἡμᾶς 
βασιλείαν. 
118 22. 9. 20. 28 39 714 142-3 1538. [Charles (ICC i. p. 15) adds 204, 
treating οἵτινες as an editorial gloss.] 
3. Col 126 τὸ μυστήριον τὸ ἀποκεκρυμμένον ἀπὸ τῶν αἰώνων καὶ ἀπὸ τῶν 
γενεῶν, νῦν δὲ ἐφανερώθη. 

Of these examples R. H. Charles rejects Jn 1535 in agreement with 
Abbott J@ 335 (“‘ the meaning is ‘it abode once for all,” i.e. aor. 
ind. in contrast with pres. ptcp.). J. H. Moulton disallows Col 153, 
accepting the punctuation in WH, and Burney dismisses Rev 118 204 
on the ground that the Hebrew construction requires that the finite verb 
should express the proper sequence of the ptcp., and not describe an event 
actually prior in time to its antecedent. 

The extreme frequency of this construction in Rev marks it as ἃ 
Hebraism there. Moreover, Dr. Charles has strengthened his case by 
giving instances in which the ptcp. is in an oblique case. 


4. Heb 8!° 10:5 (both LXX) are not pressed, for reasons given by Charles 
Studies in Apoc. 90 n.1. 


(It is quite possible that a similar Aramaic construction lies behind 
the awkward Greek in Lk 108 καὶ εἰς ἣν ἂν πόλιν εἰσέρχησθε καὶ δέχωνται 
ὑμᾶς, ἐσθίετε. . . . R. H. Charles observes the loose construction in 
1 Co 78 γυνή ἥτις ἔχει ἄνδρα ἄπιστον καὶ οὗτος συνευδοκεῖ (-- καὶ 
συνευδοκοῦντα) and suggests an idiomatic Hebrew background. But 
this could be paralleled in class. Gr. from Homer downwards, e.g. Od. 
ii. 114. See Monro Hom. Gr. 247, K-G ii. 432 f.) 

Two papyrus examples may be given. BGU iii. 846 (ii/a.p.) 
Ἤκουσα παρὰ το[ῦ υἱ]οῦ μου τὸν εὑρόντα σαι ἐν τῷ ᾿Αρσαινοείτῃ καὶ ἀκαίρως 
πάντα σοι διήγηται. Ῥ Ryl ii. 1534° (a.p. 138-161) “ If anything happen 
to my son being childless and intestate,” ἢ καὶ τέκνα μὲν εἴχοντι ἐπι- 
μεταλ(λ)άξη δὲ καὶ τὰ τέκνα “‘ or if he has children, in the case of the 
decease of those children...’ [Pallis renders Lk 108 literally.— 
R. McK.] 


1 The construction illustrated by these examples from class. Greek is not 
close enough to that found repeatedly in Rev to discount Hebraism in the 
instances cited below, 


430 A GRAMMAR OF NEW TESTAMENT GREEK. 


(>) Infinitive. 


The solitary instance of the resolution of the infinitive into a finite 
verb in the following clause is claimed by Charles for 


2. Rev 13% καὶ ἐδύθη αὐτῷ δοῦναι πνεῦμα τῇ εἰκόνι. . . καὶ ποιήσῃ, 


on the ground that the sense demands this co-ordination of ποιήσῃ with 
δοῦναι rather than with the intervening iva λαλήσῃ ἡ εἰκών. 

Cf. Tob 2: δὲ καὶ... ἔθηκα μέχρι τοῦ τὸν ἥλιον Svew καὶ θάψω αὐτόν (om. 
ΒΑ, -rw P Oxy). 0. Ο. Simpson OA i. 206. : “ An instance of the resolution 
of an infin. into a finite verb in Hebraistic style, and of its rejection or 
modification in a subsequent recension.” 


B. THE SEVERAL PARTS OF SPEECH. 
1. Pronouns. 
(a) Definite Article. 


(a) Unusual insertion of the article. ‘‘ Peculiar to Hebrew (cf. analogous 
examples in Biblical Aramaic, Dan 215 3? e¢ al.) is the employment of 
the article to denote a single person or thing (primarily one which is as 
yet unknown, and therefore not capable of being defined) as being present 
to the mind under given circumstances. In such cases in English the 
indef. art. is mostly used.” Ges-K Heb. Gr.*§ § 126, g. Wellhausen’s 
examples (W1 26, 719) support his assertion that Codex D preserves 
many Semitisms which have been pruned away in other MSS (see Proleg. 
242). With one exception, Mk 10?° διὰ τῆς τρυμαλίας τῆς papidos (where 
Mt and Lk have dropped the def. art. before each genitive), all the 
examples are peculiar to D, viz. Mk 8356 τὸ τέλος, 81} τὸ σημεῖον, 9385 τὸ 
παιδίον, 125 τοῖς γεωργοῖς, Mt 1039 τοῦ ἀσσαρίου, 144 ἐπὶ τῷ πίνακι. 


See Proleg.? 81. 236. 


(8) Omisston of the article (W1 26, 311), see Proleg.® 81. 236. 

1. Blass? 150 called attention to illustrations of the Semitic rule which 
drops the art. with a noun in construct state, and quoted many examples, 
especially from the Hebraic songs in Lk. In prepositional phrases 
Debrunner (BI-D § 259) recognises Greek usage. Wellhausen notices 
the omission of article with noun in construct state preceding a definite 
noun, Mt 12% (=Lk 1151), and compares two relics of this construction 
in Ὁ, Mt 10% εἰρήνη ὑμῶν, Lk 1119 υἱοὶ ὑμῶν, emphasising “ the extra- 
ordinary importance of this unpretentious Semitism.” 

In view of this, W. C. Allen’s claim that Mk 378 τοῖς υἱοῖς τῶν ἀνθρώπων 
is a pure Aramaism seems questionable (Exp 7' xiii. 330. See also his 
Comm. on Mk. p. 50, “1. vi. τ. ἀνθ. Ξε) 3. Cf. Dn 238 (6), where 
LXX substitutes ἀνθρώπων for oi υἱοὶ τῶν ἀνθρώπων ᾽). But see below, 
p. 441. 

The evidence of D in this matter is of dubious value considering 
the long list of omissions and additions of the article furnished by Von 
Soden (Die Schriften des NT, τ. ii. 1809), who comments on the remark- 


SEMITISMS IN THE NEW TESTAMENT. 431 


able uncertainty of its treatment of the article and attributes the 
omissions partly to scribal blunders. 
2. Torrey (H7'R xvi. 323) finds “traces” of this Semitism in six 
phrases in John: 
Jn 149 σὺ εἶ ὁ vids τοῦ θεοῦ, σὺ βασιλεὺς εἶ τοῦ ᾿Ισραήλ (for ὁ βασιλεύς). 
45 ἦν δὲ ἐκεῖ πηγὴ τοῦ Ἰακώβ. 
5°” υἱὸς ἀνθρώπου. 
5° εἰς ἀνάστασιν ζωῆς . . . εἰς ἀνάστασιν κρίσεως. 
068 ῥήματα ζωῆς αἰωνίου ἔχεις. 
9° ὅταν ἐν τῷ κόσμῳ ὦ, φῶς εἰμι τοῦ κόσμου (contr. 17 812). 
For the qualitative force of the anarthrous nouns in the last two 
exx., see Proleg. 82. 


(6) Personal Pronouns. 

(a) Redundant use to strengthen definition of noun (W 27, 219). 

1. Mk 5" Ὁ αὐτῷ τῷ δαιμονιζομένῳ. 

67 αὐτὸς γὰρ Ἡρώδης. 
618 D αὐτὴν γυναῖκα τοῦ ἀδελφοῦ σου. 
653 AC αὐτῆς τῆς Ἡρωδιάδος. 
Mt 12% D αὐτοῦ τοῦ ἀνθρώπου ἐκείνου (but not in || Lk 1155). 

In Mt 85 αὐτὸς δὲ 6 Ἰωάνης may mean “ἡ John in his person,” or ~ As 
to himself, John...” [Perhaps simply “‘ John.” So Med. and MGr. 
—R. McK. ] 

In Mk 6” W. C. Allen suggests that whether αὐτῆς or αὐτοῦ (RN BDL) 
be the original reading, the word is due to mistranslation of MM12 


ΝΥ ΤΥ ΠῚ or of NIT N3, the daughter of Herodias (op. cit. 330, also 


Mark, in loc.). In the former explanation 7 the sign of the genitive 

has been confused with the Aramaic demonstrative pronoun. (Cf. 
Stevenson Aram. Gr. pp. 24 and 18). 

2. Jn 918 τοὺς γονεῖς αὐτοῦ τοῦ ἀναβλέψαντος. 

13 ἄγουσιν αὐτὸν πρὸς τοὺς Φ. τόν ποτε τυφλόν. (Burney Aram. 

Orig. 85, who quotes Pal. Syr. in support of these 

Aramaisms, adduces 91° as a parallel to Mk 6532, and 91% 

as reproducing “‘ another peculiarly Aram. idiom,” viz. 

“the anticipation of the direct object of a verb by a 
pronominal suffix.’’) 


(8) Unusual frequency of the oblique case of the unemphatic personal 
pronoun, 6.5. of μαθηταὶ αὐτοῦ, δύο ἐκ τῶν μαθητῶν αὐτοῦ (for class. of μαθηταί, 
δύο μαθηταί). [Idiomatic in MGr.—R. ΜΟΚ.] 

Wellhausen (W1 29, 322) was here following Blass (§ 48, 2), who found 
the reason for this in the Semitic usage where these pronouns are easily 
attached as suffixes to substantival and verbal forms. But the evidence 
from the Κοινή given in Proleg. 85 is accepted by Debrunner (BI-D 
§ 278) as partial explanation. In addition to papyrus examples given in 
Proleg. and Vocab. 94, we may add P Iand 9?° (ii/a.D.) σοῦ τὸ πορφυρίιν 
ἐπί σε al vé [πεμψα], on which editor remarks, “‘ persaepe pronominis 


432 A GRAMMAR OF NEW TESTAMENT GREEK. 


genetivus sic collocatur, velut P Oxy vii. 1064° (iii/A.D.) εἰδώς σου 
τὸ σπουδαῖον. Seiungitur etiam a substantivo uno pluribusve verbis. 
velut BGU ii. 52318 oiSas μου yap τὴν πρὸς ἐμὲ καὶ σὲ diay.” 

(y) Confusion of personal and demonstrative pronouns. 


It is hardly necessary, with Wellhausen (130, 223), to posit a Semitic 
identity of pronouns as in any way the cause of a Lucan peculiarity. 
That Luke writes, 1074 ἐν αὐτῇ τῇ ὥρᾳ, where Mt 11* has ἐν ἐκείνῳ τῷ 
καιρῷ, or again in 12 when Mt 1019 and Mk 13" have ἐκείνῃ, can hardly 
be accounted for by fidelity to an Aramaic original, when we observe 
this use of αὐτὸς ὁ by Luke when Semitic sources are not in question, 
eg. Acts 1618 2213. Other examples of this mannerism are Lk 238 721 
2019. For distribution of αὐτὸς ὁ in NT see Hawkins HS? 16. For 
Hellenistic parallels to the Lucan idiom see Proleg. 91 and Vocab. 94. 

(δ) Burney (ut supr. 80 ff.) accounts for the great frequency of the 
unemphatic use of the personal pronouns in the nominative in John 
by the Semitic idiom by which the pronoun marks the subject of the 
participle. This may well be a “ secondary Semitism’”’ in John, as in 
the LXX. 


(c) Reflexive Pronouns. 


The attempt (W+ 30, 223) to find the general want of these pronouns 
reflected in the Gospels is sufficiently discussed in Proleg. 87. To what is 
said there about substitutes for the reflexive pronoun we may add 
that Mk 2° shows the equivalence of διαλογίζεσθαι ἐν ἑαυτοῖς, and dian. 
ἐν ταῖς καρδίαις ὑμῶν, which is a Semitism of vocabulary to be expected 
in sayings of Jesus or in passages which echo the language of the Old 
Testament. 

The substitution of ἑαυτούς for ἀλλήλους in D (e.g. Lk 2415. 17. 32) proves 
nothing, for it is not constant, and the reciprocal use of the reflexive 
pronoun is, apart from NT usage, common in papyri; e.g. BGU iv. 
11014 (i/B.c.) ἐπεὶ συνόντες ἑατοῖς &r[n . . .] ἐχωρίσθημεν ἀπ᾽ ἀλλήλων, 
1110! συνχωροῦμεν πρὸς ἑατοὺς ἐπὶ τοῖσδε, and so 115714 (i/B.c.). 


(d) Indefinite Pronouns. 


A Semitic origin is claimed (W1! 27, 220) for three substitutes for τις, 
where the indef. art. would be used in English. 

(a) Εἷς. 

1. From Mt, Hawkins quotes (HS? 30) 819 1874(?) 2119, 26° (“* Per- 
haps also 918, which, if εἷς is the right reading, would correspond to 
eis τῶν in Mk 5* as 2653 does to pia τῶν in Mk 14%.” The difficulty of 
εἰσελθών is shown by the corrections προσελθών NPB, τις προσελθών 
LG). To these add Mk 1017 1245 1447 (x AL), Lk 5° Ὁ (εἰς ἐν πλοῖον). 
Torrey (CDA 7) would add Ac 1210 (here improbable). 

2. Beye. 915 185: 


Radermacher (Gr.? 76 n.3) cites Strabo (p. 230) ἐπηγγείλατο ἕνα ἀγῶνα 


SEMITISMS IN THE NEW TESTAMENT. 433 


ἱππικόν. For eis c. part. gen. (e.g. Lk 512. 17 1515)=7us see Prol. 96 f., 
Vocab. 187, where papyrus evidence is supplied. See also BI-D ὃ 247 
who denies weakening in classical exx. In MGr évas is indef. art., 
Thumb Handb. 328. 


(8) ἄνθρωπος, corresponding to Aram. 2, which, in the st. abs. 
is used for guidam, and is put before and after nouns. This, of course, 
is also very common in Hebrew, and is sometimes reproduced in LXX, 
e.g. 1 ΚΙ1710 ndN TWN = γυνὴ χήρα (whence Lk 435). 

Thackeray (Gr. 45) observes this use of ἀνήρ and ἄνθρωπος, and while 
noting a similar use in Aristoph. regards it as Hebraism in OT. 

Possible examples of ἄνθρωπος with another noun in this indef. sense 
are: 


1. Mt 1119 ( =k 724) 1'328- 45 J) 52 1825 901 9138 992. 
Simple ἄνθρωπος --τις. 


9: 1 001: 720 1155: 


But Epict. iii. 23. 15 is quite parallel to this Pauline use. (For 
Greek usage see new LA, s.vv. ἀνήρ, ἄνθρωπος ; Vocab. 44. 


(y) The plur. of indef. pron. often expressed by ἀπό or ἐκ ὁ. gen. 
(=Hebr. and Aram. 13), e.g. Mk 5°” 6%, Mt 2334 (=Lk 2116). 

But in Κοινή ἀπό and ἐκ ὁ. gen. had largely replaced part. gen. (Proleg. 
72. 102, P Iand 85 note), and even in class. Gk. the part. gen. was not 
unknown as subj. or obj. of a verb, e.g. Xen. Hell. tv. ii. 20, HedAnveis 
δὲ κατὰ Θεσπιᾶς γενόμενοι ἐμάχοντό τε Kal ἐν χώρᾳ ἔπιπτον ἑκατέρων. See 
further WM 253, Brug.-Th. 442, Jannaris ὃ 1313, BI-D ὃ 164. Buck ᾿ 
(Gr. Dial. 195) remarks that part. gen. as subj. is found in Av., Lith., 
and once in Umbrian. 


(δ) The negative of the indef. pron. assumes some unusual forms. 

(i) Eis... ov. Mt 1029 ἕν ἐξ αὐτῶν οὐ πεσεῖται is claimed as 
“unadulterated Semitism”’ (W! 31, 724). It may be an extension of 
the usage found in Mt 518, Lk 114°, for which parallels in Dem. c. Onet. i. 33, 
Xen. Anab. v. vi. 12, are given in BI-D ὃ 302. [Med. Gr.—R. McK.] 

The emphatic οὐδὲ... εἷς is sufficiently attested in classical and 
Hellenistic Gk., and οὐδείς ἐστιν ὅς is acknowledged (W? 24) to be 
“certainly not unGreek,”’ though Wellhausen compares it with Syr. 
lait de. 

(ii) πᾶς... ov, for Hebrew and Aramaic G5. ae «ὃ. 

To what has been written by Moulton in CR xv. 442, add reff. given 
above, p. 22 n.3. Ὁ. 8. Sharp cites Epict. ili. 22. 36, πᾶσα ψυχὴ ἄκουσα 
στέρεται τῆς ἀληθείας. R. Law (Tests of Life, 379), commenting on 
1 Jn 219, ** It seems questionable whether this is a Hebraism, as is usually 
said. The explanation of the idiom probably is, not that πᾶς was used 
in a consciously distributive sense, but that, in vernacular Greek, the 
negative was attached in sense to the verb, where we attach it to the 
nominative (‘all are-not’=‘ none are’). The attachment of οὐ to what 
seems to us the wrong word is not unusual in Greek [e.g. in Aristoph. 


434 A GRAMMAR OF NEW TESTAMENT GREEK. 


Vesp. 1091, πάντα μὴ δεδοικέναι Ξε: μηδὲν dedocxéeva.—J. H. M.J, and is 
invariable in the common οὔ φημι τοῦτο etvac=I say that this is not so.” 
[Rare in Med. Gr.—R. McK. ] 


1. Mk 132° (=Mt 2433) οὐκ ἂν ἐσώθη πᾶσα σάρξ. 

Lk 137 οὐκ ἀδυνατήσει παρὰ τ. θεοῦ πᾶν ῥῆμα. (Not a quot. from 
LXX or Heb. of Gen 1813.) 

Ac 1014 οὐδέποτε ἔφαγον πᾶν κοινόν. 

2. ὅπ 689 ἵνα πᾶν ὃ δέδωκέν μοι μὴ ἀπολέσω ἐξ αὐτοῦ. (See also 
p. 424.) 
112° πᾶς ὁ ζῶν καὶ πιστεύων εἰς ἐμὲ οὐ μὴ ἀποθάνῃ εἰς τὸν αἰῶνα. 
1246 ἵνα πᾶς ὁ πιστεύων εἰς ἐμὲ ἐν τῇ σκοτίᾳ μὴ μείνῃ. 

1 Jn 22! πᾶν ψεῦδος ἐκ τῆς ἀληθείας οὐκ ἔστιν. (In the similar 
construction in 2": 38. 3% ® 48 518 the πᾶς is positive, and 
the οὐ negatives the verb.) 

Rev 716 οὐδὲ μὴ πέσῃ ἐπ᾽ αὐτοὺς ὁ ἥλιος οὐδὲ πᾶν καῦμα. 

1822 καὶ πᾶς τεχνίτης πάσης τέχνης οὐ μὴ εὑρεθῇ ἐν σοὶ ἔτι. 
2121 καὶ οὐ μὴ εἰσέλθη εἰς αὐτὴν πᾶν κοινόν. 
223 καὶ πᾶν κατάθεμα οὐκ ἔσται ἔτι. 


(Charles has not included this construction in his list of Hebraisms.) 


8. Rom 32°=Gal 215 (=Ps 143? LXX for: ΠΡῸΣ 2B) PAY NP 3). 
Eph 4539 πᾶς λόγος σαπρὸς ἐκ τ. στόματος ὑμῶν μὴ ἐκπορευέσθω. 
5° ὅτι πᾶς πόρνος . . . οὐκ ἔχει κληρονομίαν. 
4. 2 Pet 130 ὅτι πᾶσα προφητεία γραφῆς ἰδίας ἐπιλύσεως οὐ γίνεται. 
Cf. Didache 2? οὐ μισήσεις πάντα ἄνθρωπον. Protev. Jac. vi. 1, πᾶν 
κοινὸν καὶ ἀκάθαρτον οὐκ εἴα διέρχεσθαι δι᾽ αὐτῆς. 


WM 215 observes that “Ὅ{Π1|5 Hebraism should in strictness be 
limited to the expression οὐ (μὴ) . . . mas; forin sentences with πᾶς. 
ov (μή) there is usually nothing alien to Greek usage.” 

For the latter Radermacher (Gr.? 220) cites Dion. H. Hp. ad Pomp. 
756 R, οὐκ ἀπὸ τοῦ βελτίστου πάντα περὶ αὐτῶν γράφων. Wackernagel, 
Vorlesungen ii. 274, cft. Propertius ii. 28. 18, semper, formosae, non nostis 
parcere verbis, ‘‘ niemals versteht ihr.” 


(e) The Relative Pronoun. 


(a) The Hebrew construction by which indeclinable ἽΝ is followed 


by a pronoun or pronominal suffix is paralleled in Aramaic by the use 
of 3 Or ‘4 indecl. 


1. The passages which have been claimed as examples of this 
Semitism (see Blass Gr.2 175, BI-D § 297, W1 22, #15, Burkitt Hv. d. M. 
ii. 75) are Mk 17 (=Lk 815. Note Mt corrects, also Luke in Ac 13%5), 725 
(Note 8B omit αὐτῆς). Mt 10% D, ἡ πόλις εἰς ἣν εἰσέλθητε εἰς αὐτήν, 18?° 
D, παρ᾽ οἷς οὐκ εἰμὶ ἐν μέσῳ αὐτῶν. Lk 812 D, ὧν ἔρχεται ὁ διάβολος καὶ αἴρει 
ἀπὸ τῆς καρδίας αὐτῶν τὸν λόγον. 

Mt 312 (=Lk 31’) is normal Greek, as Burney Aram. Orig. 85 n. seems 
to allow. 


SEMITISMS IN THE NEW TESTAMENT. 435 


The conjecture that Lk 104+ should read, in the absence of all textual 
evidence, ἧς (for ἥτις) οὐκ ἀφαιρεθήσεται αὐτῆς (=she has chosen the better 
part from which she shall not be taken away) was offered in W! 22, but is 
withdrawn in the 2nd ed. 

Of the same kind is Mk 1319 θλίψις, οἵα οὐ γέγονεν τοιαύτη (N.B.—Mt 
corrects, θλίψις μεγάλη, οἵα οὐκ ἐγένετο). Hawkins HS? 134 points out 
that this does not occur “in Dn 12! (either LXX or Theod.), which is 
here being referred to. See, however, Gen 41'9; and compare ἥτις τοιαύτη 
in Ex 974 and 118. Somewhat similar is Mark’s ofa . . . οὕτως in the 
best texts of 93.” 

Cf. Ac 1517 (LXX). 


2. Burney (Aram. Orig. 85) cites Jn 157 133. 1376 189 (also 955, see (3) 

below. 18° is doubtful). 
In the Apocalypse Moulton notes six examples: Rev 88 72" 

138-12 208 (to which Charles adds, ὅπου... ἐκεῖ 12°44, and 
ὅπου ... ἐπ᾽ αὐτῶν 179). 

3. Philem 12 is not an instance, since αὐτόν is emphatic, with the 
following clause in apposition. 

4. 1 Pet 272 8*LP οὗ τῷ μώλωπι αὐτοῦ. (See Prol.? 237.) 


Moulton discussed the question in Proleg.* 94 f., 237, 249. In Hinl. 
150 f. he cites, after Helbing (Gr. p. iv), P Oxy i. 117 (ii/iii A.D.) ἐξ ὧν 
δώσεις τοῖς παιδίοις σου ἕν ἐξ αὐτῶν, but quotes Wackernagel (ThLZ xxxiv. 
227) as thinking that the equivalence of MGr ποῦ clauses and such 
sentences as ois ἐδόθη αὐτοῖς is not proved. (See Psichari, 182 f.). 

Thackeray (Gr. 46) finds the construction in all parts of the LXX, 
where it ‘‘ undoubtedly owes its frequency to the Hebrew original. But 
the fact that it is found in an original Greek work such as 2 Mac (1257 
ἐν ἣ - -. ἐν αὐτῇ) and a paraphrase such as 1 Esdras (35: " 452. 68. 63) is 
sufficient to warrant its presence in the Kown.” We may add that some- 
times, as in Is 11 ev 7 . . . ἐν αὐτῇ, the construction is used in the 
Greek with no corresponding use in the Hebrew (see Ottley Isaiah i. 
p- 41, ii. p. 108). 

Canon Box has shown how this Semitism has infected the Latin of 
4 Ezra: eg. 428 de quo me interrogas de eo. Cf. also 44: 614-29 1376 
(OA ii. 547). 

A common classical usage, when a relative clause is continued by a 
clause co-ordinate with it, is to abandon the relative construction in the 
second clause and to replace the relative by a personal or demonstrative 
pronoun (cf. Xen. Cyrop. mt. i. 38, Iv. i. 15, v. ii. 15). Mr. G. R. Driver 
(Orig. Lang. 4) quotes, for the extension of this idiom “ to single-limbed 
relative clauses,” Soph. Phil. 315 (MSS), ois ᾿Ολύμπιοι δοῖέν ποτ᾽ αὐτοῖς. 
Callim. Hpigr. 43, ὧν ὁ μὲν αὐτῶν. Anth. Pal. vii. 72, ὧν ὁ μὲν ὑμῶν. For 
further exx. see Radermacher Gr.? 217, and Jannaris ὃ 1439. We may 
add Clem. ad Cor. 21° οὗ ἡ πνοὴ αὐτοῦ ἐν ἡμῖν ἐστίν. [Med. Gr.—R. McK. ] 


(8) The same particle 4 (1) can also introduce a subordinate clause 


and may be rendered by ὅτι, or iva. (See below, pp. 469 f.) 


436 A GRAMMAR OF NEW TESTAMENT GREEK. 


1. It has been suggested that sometimes these particles in the Greek 
text mistranslate the Aramaic relative. Thus W. C. Allen (Hxp T xiii. 330 
and Comm. in loc.) explains Mk 8*4 ὅτι ὡς δένδρα ὁρῶ περιπατοῦντας, where 
ὅτι--- Ξεοὕς. So W! 22, 215, explains Mk 4% οὐ γάρ ἐστιν κρυπτόν, εἰ μὴ 
ἵνα φανερωθῇ (=nisi quod reveletur=quod non reveletur). The parallels 
in Mt 1076 and Lk 12? (ὃ οὐκ ἀποκαλυφθήσεται), and Lk 817 (ὃ οὐ μὴ γνωσθῇ 
καὶ εἰς φανερὸν ἔλθῃ) support Wellhausen. 

We may, however, cite Epict. Ench. 51, ποῖον οὖν ἔτι διδάσκαλον 
προσδοκᾷς, ἵνα εἰς ἐκεῖνον ὑπερθῇ τὴν ἐπανόρθωσιν ποιῆσαι τὴν σεαυτοῦ; 
where Melcher (De Sermone Epicteteo 85) observes “ Att. εἰς ὅντινα ὑπερ- 
Anon.” 

So Epict. i. 24, 3, οὐδεὶς δὲ δειλὸν κατάσκοπον πέμπει, ἵν᾽, ἂν μόνον 
ἀκούσῃ ψόφου καὶ σκιάν ποθεν ἴδῃ, τρέχων ἔλθῃ τεταραγμένος... (Att.: 
ὅστις . . . ἐλεύσεται, Vel πρόσεισιν .. .). 

Moulton (Hinl. 332 n.) accepted Wellbhausen’s explanation of Mk 45 
(retained in Mt 8”, slightly changed in Lk 8535), ris ἄρα οὗτός ἐστιν, ὅτι 
καὶ ὁ ἄνεμος καὶ ἡ θάλασσα ὑπακούει αὐτῷ ; where ὅτι is used only to avoid 
ᾧ ... αὐτῷ. Lagrange, however (S. Mare, p- xc), cites Plato Huthyph. 
2a: τί νεώτερον, ὦ Σώκρατες, γέγονεν, ὅτι σὺ... διατρίβεις ; 

There seems less reason to follow Wellnacan in taking ὅτιτε ὅς in 
Mt 112%, or in reversing the process in Mt 1110 (“‘ still more than a prophet 
is this, for about him (Ὁ Dy) is it said’’). 


2. Burney (Aram. Orig. 75 ἴ., 101 ff.) discovers many such mistransla- 
tions, supporting his contention in some instances by quoting the Syriac 
or Arabic versions : 


iva for relative: 

Jn 18 57 639. 50 986 1416. 

Rev 191° is quite parallel to Jn 6°°. 
ὅτι for relative : 

Jn 845 917 (Ὁ 138), 

N.B.—The converse is suspected by Burney (2b. 29, 34) in Jn 15: 15, 
with Torrey’s strong endorsement (H7'R xvi. 328) : 


Jn 14 punctuating ὃ γέγονεν ἐν αὐτῷ ζωὴ ἦν, and taking ὃ γέγονεν -- 
Nu, the result is “‘ inasmuch as in Him was life.” 

113 supporting the poorly attested qui natus est by showing that 

in Aramaic the final ἢ alone distinguishes the sing. verb from 

the plur., whilst the next sentence begins with Ὶ (καί). “‘ Who 


39 


believe on His name, inasmuch as He was born, not. . . 


In neither instance is there any necessity for this explanation of the 
relative pronoun. 

Mr. G. R. Driver (ut swpr. 3) remarks (a) that if this usage is due to 
Semitic influence, it is strange that the Hebr. WX is never so rendered 
in the LXX; (6) in no case in Jn is this translation necessary ; (c) in 
every case (exc. 18 and 15) the constr. occurs in words presumably 
spoken originally in Aramaic. The last observation tells against Burney’s 


SEMITISMS IN THE NEW TESTAMENT. 437 


theory that Jn was (mis-)translated from a written Aramaic original, 
but not against the possibility that we have here a genuine Aramaism. 

Lagrange (S. Jean, p. cix) shows that in the Ancyra inscr. (Cagnat ili. 
188) the Lat. ex quo . . . darentur is rendered, ἵνα ἐξ αὐτοῦ. . . δίδωνται, 
whilst in two other passages iva represents ut. This, coupled with the 
examples given above from Epictetus, shows that by this time iva was 
used in Greek as equivalent to a relative. We may therefore speak of 
this use in Jn as a secondary Semitism. 


(y) The indeclinable particle 4 is suspected by Burney (ué supr. 


101 ff.) to lie behind the difficult syntax of the relative pronoun in this 
characteristic group of passages in the Fourth Gospel. 


2. Jn 1059 ὁ πατήρ μου 6 δέδωκέν μοι πάντων μείζων ἐστιν. (6 SB*LW, 
ὃς Α. μείζων NLW, μεῖζον AB). 

1711: τήρησον αὐτοὺς ἐν τῷ ὀνόματί σου ᾧ δέδωκάς μοι. 

eyo ἐτήρουν αὐτοὺς ἐν τῷ ὀνόματί σου ᾧ δέδωκάς μοι. 
(ᾧ SABCLW, οὖς D*, ὅ D*). 
Cf. 113: Πατήρ, ὃ δέδωκάς μοι, θέλω ἵνα ὅπου εἰμὶ ἐγὼ κἀκεῖνοι ὦσιν 
μετ᾽ ἐμοῦ. 

17? ἵνα πᾶν ὃ δέδωκας αὐτῷ δώσει αὐτοῖς ζωὴν αἰώνιον. Burney 
suggests πᾶν ὅ-- Ἢ NB Aramaic for “‘ all who,” “ every 
one who,” “‘ all which”’ ; so Hebr. {bz ““ the whole of it,” 
with plur. reference, ef. Ex 14’. 

6%" πᾶν ὃ δίδωσίν μοι ὁ πατὴρ πρὸς ἐμὲ ἥξει. 

39 ἵνα πᾶν ὃ δέδωκέν μοι μὴ ἀπολέσω ἐξ αὐτοῦ, ἀλλὰ ἀναστήσω 

αὐτὸ τῇ ἐσχάτῃ ἡμέρᾳ. ! 

Cf. 1 Jn δ᾽ ὅτι πᾶν τὸ γεγεννημένον ἐκ τοῦ θεοῦ νικᾷ τὸν κόσμον, which 
shows that the neut. πᾶν is used as a collective pronoun. 


In none of these instances does Lagrange so much as suggest Aramaic ; 
the attraction of the relative is offered as the explanation of 1711: 12, 
Mr. G. R. Driver does not dispute the Aramaic origin of the idiom, but 
observes that in every case the passage is attributed to Jesus, and is 
not evidence of an Aramaic Gospel translated into Greek, but of the 
Aramaic of the ipsissima verba of our Lord. 


(δ) By this ambiguous use of 1 Burney accounts for the Lucan varia- 


tion oi βλέποντες for the Matthaean ὅτι βλέπουσιν... ὅτι ἀκούουσιν, .. 
Pyows PIAA in Mt 136 17—Lk 10-4 (The Poetry of our Lord, 145). 


(f) Distributive Pronouns and Pronominal Adjectives. 


(a) The absence in Hebrew and Aramaic of special words correspond- 
ing directly to ἄλλος or ἕτερος, involves the use of certain Semitic idioms 
to express the idea alter . . . alter. (See G-K § 139 (c).) 

These are : 

(1) WS with YON or YI as correlate. Gen 13 (LXX ἕκαστος 


ἀπὸ τοῦ ἀδελφοῦ αὐτοῦ). 


438 A GRAMMAR OF NEW TESTAMENT GRERFK. 


(2) ΠῚ... ΠῚ Ex 14° (LXX καὶ οὐ συνέμιξαν ἀλλήλοις), Is 68 (LXX 
ἕτερος πρὸς τὸν ἕτερον). 

(3) THN... THN 2 Sam 1465 (LXX καὶ ἔπαισεν ὁ εἷς τὸν ἕνα 
ἀδελφὸν αὐτοῦ). 

(4) The substantive repeated Gen 4753: (LXX ἀπ᾽ ἄκρων ὁρίων Δἰγύπτου 
ἕως τῶν ἄκρων), cf. Dt 4°? 28%, 


(2) and (3) are both found in 1 Sam 144 [12 eee my (LXX ἔνθεν 
. ἔνθεν) and TANT 3 a5 TANT (LXX τῷ ἑνὶ... τῷ ἄλλῳ). 


Thackeray (Gr. 45) observes: ‘‘ The rarity of phrases like ἕτερος τὸν 
erepov (still found in the Pentateuch, Isaiah and the early chapters of 
Kzekiel) is partly due to the tendency in the Kowy to abandon words 
expressive of duality. But it is noticeable that the use of ἀνήρ Ξεἕκαστος 
is practically confined to one group of books,’ in which “ ἕκαστος, 
which is freely used in other parts of the LXX, is either wholly or nearly 
unrepresented.” 


1. In the Synoptic Gospels (1) and (2) are not represented, but the 
idiom of (3) may possibly be found in Mk 1527, Mt 2021 2440% 9738, 
Lk 1810 D, eis Φαρισαῖος καὶ εἷς τελωλωνὴς (sic), and (4) in Lk 11}? οἶκος 
ἐπὶ οἶκον πίπτει, Mk 13? (=Mt 242=Lk 218) λίθος ἐπὶ λίθῳ, Mt 23% ἀπὸ 
πόλεως εἰς πόλιν. (W?! 30, 223.) “‘ From city to city,” however, is quite 
idiomatic English and is not necessarily Hebraic. A closer parallel than 
this seems to be Lk 1753: ὥσπερ yap ἡ ἀστραπὴ ἀστράπτουσα ἐκ τῆς ὑπὸ τὸν 
οὐρανὸν εἰς τὴν ὑπ᾽ οὐρανὸν λάμπει. 

2. (3) is found in Jn 202, 

3. and in Gal 433. 

But in both exx. eva... καὶ ἕνα is probably due to the gradual 
disappearance of μὲν... δέ in Hellenistic. 

See BI-D, § 247. 3, and for vernacular use, Vocab. 187. [(4) Med. and 
MGr.—R. McK.] 

(8) Closely akin to the idiom of (3) above is ] Ki 227 ὉΠ 198 7273 
Let your speech be like the speech of the rest of them. This has been cited 
(W* 30, 329) to explain Mk 6 ὅτι προφήτης ὡς cis τῶν προφητῶν, A 
prophet as another prophet, like any other prophet. Moffatt translates, 
It 1s a prophet like one of the old prophets. Lk 98 corrects, ὅτι προφήτης 
τις τῶν ἀρχαίων ἀνέστη. 

(γ) For “the rest’? as compared with a single example of a class, 
Hebrew and still more Aramaic use by. 

Thus Gen 31 “The serpent was more subtle than any other beast of the 
field,” Ten nn 550, 

Gen 43*4, “‘ And Benjamin’s mess was five times as much as any of the 
rest of theirs,” pps NNW, 

This use of πᾶς may possibly be seen (so [731 31, 223) in Mk 433, 
Lk 830. 21 132-4, though in Lk 32°21 the explanation is far-fetched. 
[MGr sometimes uses ὅλος in much the same way.—R. McK.] 


SEMITISMS IN THE NEW TESTAMENT. 439 


2. NUMERALS AND DISTRIBUTIVES. 
Semitic influence has been suspected in— 


(a) The use of Cardinals for Ordinals in dating Incidents. 
Cf. Gen 15 IM& DY (LXX, ἡμέρα μία). So in Aramaic (Kautzsch Gr. 122, 
Marti Gr.? 82, Dalman Gr.? 131). See Proleg. 95 f., 237. 
1. Cf. Mk 165, Lk 211, τῇ μιᾷ τῶν σαββάτων. 

Mt 28: εἰς μίαν caf. 

Ac 207 ἐν τῇ μιᾷ τ. σαβ. [Plummer ([CC, p. 407) suggests this 
meaning (unnecessarily) in Lk 1738 ὅτε ἐπιθυμήσετε μίαν 
τῶν ἡμερῶν τοῦ υἱοῦ τ. ἀνθρώπου ideiv.] 

2. Jn 201-19, [Charles unnecessarily suggests this in Rev 64. See 
ICC i. p. exlviii.] 
3. 1 Co 16?. 
See however p. 174, Proleg. 96, and for MGr, Thumb Handb. 82. 
[Med. Gr.—R. McK.] 


(Ὁ) The use of Cardinals for Adverbials. 


1. Mk 48: 2° εἰς τριάκοντα καὶ ev ἑξήκοντα καὶ εν ἑκατόν. 

““The MSS offer many variations and combinations of εἰς and εν. 
But whatever be original, it is no doubt due to over-scrupulous transla- 
tion of 4m,” W. C. Allen (Hzp T xiii. 330), who cites Dn 31° nyaw-In 


seven times, and Gen 26 (Targ. Onk.) AND 1h by one hundredfold. 


‘““ The writer of the First Gospel has avoided the Aramaism by substituting 
Gin τ΄. δ᾽ (Comm: Marks 79). 


(c) Distributives expressed by Repetition, either of the Cardinal 
Number or of the Noun itself.—The former is literally reproduced in 
LXX, e.g. in eis εἷς 1 Chr 245, δύο δύο Gen 61°, ἑπτὰ ἑπτά Gen 73. (This is 
also Aramaic, Dalman Gir.2 135.) The latter e.g. in 2 Chr 341° ἐργασίᾳ καὶ 
ἐργασίᾳ (= may may? , in every department of work). Sometimes 
κατά is combined with this reduplication, e.g. κατ᾽ ἐνιαυτὸν ἐνιαυτόν 1 Καὶ 
78, κατὰ μικρὸν μικρόν Dt 7533, κατὰ φυλὰς φυλάς Zech 1215, 

1. The NT supplies δύο δύο Mk 6’; συμπόσια συμπόσια, πρασιαὶ πρασιαί 
Mk 639:  δεσμὰς δεσμάς Mt 1530 Epiph. ; ἀνὰ δύο δύο Lk 10! 
Β αἱ. ; εἷς κατὰ εἷς (Mk 149) is claimed as a hybrid confusion 
between the Aram. ἽΠ ἽΠ and the vulgar Greek καθεῖς. 

3. 2 Co 416 ἀλλ᾽ ὁ ἔσω ἡμῶν [ἄνθρωπος] ἀνακαινοῦται ἡμέρᾳ καὶ ἡμέρᾳ is 
regarded as Hebraism by Bl-Debr. § 200. 1. -- 1) δ᾽" 
(contrast Heb 3:5 καθ᾽ ἐκάστην ἡμέραν). 

See Ῥχοίεσ.3 21 n.°, 97, for the significance of these locutions, with the 
important additions made by Moulton in Hinl. 156f. (See also ὃ 104 
above.) Psichari (183 ff.) adduces MGr examples to support Thumb’s denial 
of Hebraism (Hellen. 128, Handb. 83). Wackernagel (TALZ xxxiv. 227), 
however, recognises a Semitism in Mk 68°". G. and H. point out on P Oxy 


440 A GRAMMAR OF NEW TESTAMENT GREEK. 


vi. 9408 (vi/a.D.) that σου μίαν μίαν means together with you, and is not 
distributive in that passage. The new LS cites Soph. Frag. 201, μίαν 
μίαν (Ξεκατὰ μίαν). 


3. ADJECTIVES AND ADJECTIVAL SUBSTITUTES. 


In Hebrew the wide use of the construct state largely took the place 
of the adjective. Greek has many corresponding uses of the genitive 
case of a noun to mark description, material etc. The two most char- 
acteristically Semitic idioms are (1) the genitive of an abstract noun in 
place of an adjective of quality, and (2) the use of vids with a following 
genitive of origin or definition. 

(1) The so-called “‘ Hebraic Genitive’ is an extension of the con- 
struction found in Greek poetry, as shown in Proleg. 74, 235. In the 
LXX Thackeray regards it as “‘ partly but not altogether due to literal 
translation” (Gr. 23). The same may be said of the NT instances. 
To the classical instances referred to in Proleg.* (ut supra), Radermacher 
(Gr.2 109, 111) adds Herodotus iv. 136 αἵ re ἡμέραι ὑμῖν τοῦ ἀριθμου 
διοίχηνται. From late Greek he cites Demosth. In Midiam 93 (“δὴ 
interpolated document of the Hellenistic age”) ἡ κυρία τοῦ νόμου 
(the legal limit—contrast ἡ κυρία in ὃ 84, the equivalent Attic 
term); Marcellinus Vit. Thuc. 57, λόγοι εἰρωνείας (as well as λόγοι 
eipwvixot), pseudo-Chion Hp. 16. 3, ἀπεχθείας ἔργον, and, in view of 
Pauline parallels, pseudo-Hippocr. Hp. 10. 6 σῶμα σοφίης. Thumb 
(Brugmann Gr.’ 677) dissents from Wackernagel’s assumption that MGr 
ἄνθρωπος τῆς μπιστοσύνης, a trustworthy man, is a Hebraism. 


1. Mk 276 (=Mt 124=Lk 65) οἱ ἄρτοι τῆς προθέσεως a t.t. from the OT. 

Lk 453 οἱ λόγοι τῆς χάριτος, 108 οἰκονόμος τῆς ἀδικίας, ὃ μαμωνᾶς τῆς 
ἀδικίας, 185 κριτὴς τῆς ἀδικίας. 

Ac 64 δ ῥήματα βλασφημίας, 8538 χολὴ πικρίας, 915 σκεῦος ἐκλογῆς. 

2. Rev 13! 175 ὀνόματα βλασφημίας. 
3. Rom 176 πάθη ἀτιμίας, 831 τὴν ἐλευθερίαν τῆς δόξης κιτιλ., 1239 (LXX) 
ἄνθρακες πυρός. Possibly Eph 1" 4533 (see p. 485). 

Phrases with σῶμα. Rom 6°10 σ. τῆς ἁμαρτίας, 774 τὸ σ. τοῦ θανάτου, 
Ph 37 τὸ σ. τῆς ταπεινώσεως ἡμῶν, τὸ σ. τῆς δόξης αὐτοῦ, 
Col 1533 211 τὸ σ. τῆς σαρκὸς (αὐτοῦ). 

Phrases with ἡμέρα, Rom 2° ἡμ. ὀργῆς, 2 Co 6? ἡμ. σωτηρίας (LXX), cf. 
1 Pet 213 ἡμ. ἐπισκοπῆς (LUXX). These are rooted in the 
language of the OT, but they can only be termed 
Secondary Semitisms. 

4, Heb 1218 ρίζα πικρίας (LXX Dt 2918 AF; see BI-D ὃ 165). 

Jas 155 ἀκροατὴς ἐπιλησμονῆς. Perhaps νόμον τέλειον τὸν τῆς ἐλευ- 
θερίας (tb.) and τὸ πρόσωπον τῆς γενέσεως αὐτοῦ (38) come 
under the same head. 

Debrunner (BI-D § 165) includes Ac 118, 2 Pet 2%. But μισθὸς (τῆς) 


ἀδικίας is an objective genitive. (Cf. Ezek 144 ἡ κόλασις τῆς ἀδικίας αὐτοῦ, 
4415 εἰς κόλασιν ἀδικίας.) 


—— 


SEMITISMS IN THE NEW TESTAMENT. 44] 


(2) The use of vids or τέκνον with genitive in metaphorical sense. | 

Thackeray (Gr. 41) observes that in LXX “this Hebraism is mostly 
confined to the literal group: the Hexateuch, Isaiah and Chronicles 
generally avoid 10. Wellhausen sees in this a common mark of genuinely 
Aramaic style (cf. Arabic dhu), W127. NT instances are : 


1. Mk 219 (=Mt 94 =Lk 5%) vi. τ. νυμφῶνος; 3:7 vi. βροντῆς. 

Mt 23% vi. γεέννης. 

Lk 108 vi. εἰρήνης, 168 2033 of vi. τ. αἰῶνος τούτου, 208° vi. τ. 
ἀναστάσεως. 

Ac 488 υἱὸς παρακλήσεως. 

2. ὅπ 1712 ὁ vi. τ. ἀπωλείας. 
8. 1 ΤῺ 5 vi. φωτός, 2 Th 23 ὁ vi. τ. ἀπωλείας. 

Rom 98, Gal 438 τὰ τέκνα τ. ἐπαγγελίας. 

Eph 2? δϑ vi. τ. ἀπειθείας (whence imported into text of Col 3° in inferior 
MSS), 23 τέκνα ὀργῆς, 5° τέκνα φωτός, Col 118 τ. vi. τ. 
ἀγάπης αὐτοῦ. 

4. 1 Pet 1"} τέκνα ὑπακοῆς, 2 Pet 2! κατάρας τέκνα. 

Ν.8.---Οἱ υἱοὶ τ. πονηροῦ (Mt 1338), vi. διαβόλου (Ac 1310), as also τὰ 
τέκνα τοῦ διαβόλου (1 Jn 310), hardly come under this heading. 

Deissmann (BS 161) accepts these in 1 and 2 as translation-Greek. Of 
those in 3 and 4 he says, “ In no case whatever are they un-Greek ; they 
might quite well have been coined by a Greek who wished to use im- 
pressive language. Since, however, similar terms of expression are found 
in the Greek Bible, and are in part cited by Paul and others, the theory 
of analogical formations will be found a sufficient explanation.” 


(3) The phrase ὁ vids τοῦ ἀνθρώπου, a literal translation of the 
Aramaic NWIN 12, was quite unintelligible except on Palestinian soil. 
Apart from its frequent use in all the Synoptic Gospels (on the lips of 
Jesus), it is found in— 


1. Ac: 75°, 
2s Jn 152 318. 14 627. 53. 62 828 935 1228. 34 1331, 


The phrase vids ἀνθρώπου occurs 3 times. In Jn 5%’ the anarthrous 
form is probably due to the writer’s sense that the title is here used 
qualitatively. In Rev 1? 144 the wording is doubtless influenced by 
the LXX of Dn 1713, 

On the whole subject see Dalman WJ 234-267. 


4. DEGREES OF COMPARISON. 


(a) The absence of degrees of comparison in Semitic languages (other 
than Arabic) may account, according to Wellhausen, for the use of the 
positive adjective in some passages in the Gospels. (See W1 28, 221, 
and Moulton’s discussion in Hinl. 124 1.) 


1. Mk 1055. Clement of Alex. reads εὐκόλως for εὐκοπώτερον. Τάχιον (D) 
suggests an independent smoothing of εὐκόλως. 


442 A GRAMMAR OF NEW TESTAMENT GREEK. 


Mt 22% ποία ἐντολὴ μεγάλη, a less idiomatic rendering of the Aramaic 
(from Q?) than πρώτη πάντων in Mk 1228. Note that Mt 
2288 μεγάλη καὶ πρώτη ἐντολή is quoted by Justin M. Ap. 
1. 16 as μεγίστη. 

Lk 5°° ὁ παλαιὸς χρηστός ἐστιν, according to Wellhausen (Das Ev. 
Iuc. 19), must be taken as comp. or superl. Plummer 
(who regards xpyordrepos, AC vg, as a corruption) and 
Lagrange defend the positive interpretation. 

Mk 948: 49-47 καλόν... 7, whilst in 4? μᾶλλον is inserted. 


(5) The comparative particle is sometimes used after a verb as though 
by itself it meant “‘ more than.” 


1. Mk 3* Lk 157. 
Lk 17? λυσιτελεῖ αὐτῷ... ἢ... (contrast Mt 18° cippeper αὐτῳ 
iva, Mk 9% καλόν ἐστιν αὐτῷ μᾶλλον et... ). 


W? 21 acknowledges that this is found in Plut. 
For classical instances see WM 302, Riddell, Plat. Apol. p. 183, K-G ii. 
303. [Rare in Med. and MGr.—R. McK.] 


3. 1 Co 1419 θέλω... ἤ. Cf. the agraphon in Justin M. Ap. i. 15, θέλει 
yap ὁ πατὴρ ὁ οὐράνιος τὴν μετάνοιαν τοῦ ἁμαρτωλοῦ ἢ THY 
κόλασιν αὐτοῦ. Gildersleeve (in loc.), cft. Eur. Tel. fr. 
714 N?: 

σμίκρ᾽ ἂν θέλοιμι καὶ καθ᾽ ἡμέραν ἔχων 
ἄλυπον οἰκεῖν βίοτον ἢ πλουτῶν νοσεῖν. 

(c) For reduplication to express the elative force of the adjective, see 
Moulton’s treatment, § 104 above; Delbriick Grd. v. (iii.) 139 ff. Wetstein 
(ap. Heb 1037), cft. Aristoph. Vesp. 213 τί οὐκ ἀπεκοιμήθησαν ὅσον ὅσον 
στίλην; For numerous parallels see Radermacher Gr.? 68 n.t. [MGr. 


—R. McK.] 


1. Lk 5° Ὁ ὅσον ὅσον for ὀλίγον. 
4, Heb 10%’, which may be an echo of Is 267°, μικρὸν ὅσον ὅσον, in 
introducing the citation from Hab 2*-. 

Conybeare and Stock (Selections from LXX, 77) refer to σφόδρα 
σφόδρα, Ex 17: 15, Num 14’, Ezek 9°, Jdth 42; σφόδρα σφοδρῶς Gen 719, 
Jos 3%; θιμωνιὰς θιμωνιάς Ex 84, ἄνω avo... κάτω κάτω Dt 28%. 
They observe, “‘ In all the above instances the kind of intensification 
involved is that of a repeated process.” 


(4) For the comparative use of παρά see below, under B 8. 


(6) The superlative idea was sometimes expressed in Hebrew by 
adding oviod to the adjective. This could be rendered literally in the 


LXX by the “ dative of the person judging.” Thus Jon 88, πόλις μεγάλη 


τ Mr. E. E. Genner tells me that this passage is cited by two authorities, 
one of which gives μᾶλλον instead of βίοτον. 


᾿. «(ἡ 


SEMITISMS IN THE NEW TESTAMENT. 443 


τῷ θεῷ, an exceedingly great city. Two possible instances are found in 
NT: 


1. Ac 729 ἀστεῖος τῶ θεῷ, exceedingly fair. 

8. 2 Co 10% δυνατὰ τῷ θεῷ divinely strong (Moffatt). In this passage 
more probably dat. commodi. For the former see Prol. 
104, also p. 166 above. 


(f) A well-known Hebrew equivalent for the superlative Dw wap 
has sometimes been discovered in— 


2. Rev 1918 βασιλεὺς βασιλέων, κύριος κυρίων. 
3. 1 Tim 615 ὁ βασιλεὺς τῶν βασιλευόντων καὶ κύριος τῶν κυριευόντων. 
4, Heb 98 σκηνὴ ἡ λεγομένη ἅγια ἁγίων. 

The last is Hebraic, but is introduced as a stereotyped t.t. In the 
other two, Bac. Bac. means “ruler over kings.” Cf. P Leid. W aly. 9 
ἅγιε ἁγίων. P Par 51*4 (=UPZ i. p. 360, ii/B.C.) ἐλθέ μοι θεὰ θεῶν. For 
MGr, Thumb (Handb. 33) gives γεναῖκα τῶν γεναικῶν “ a queenly woman,” 
σκλάβος τῆς σκλαβιᾶς “‘ a vile slave.” 


5. ADVERBS AND ADVERBIAL LOCUTIONS. 


** Adverbs derived from adjectives are certainly common in later 
Syriac, but are not properly Semitic” (W? 28, 5321). 

Various substitutes are employed : 

(1) In Hebrew “‘ the infin. absol. occurs most frequently in immediate 
connexion with the finite verb of the same stem, in order in various ways 
to define more accurately or to strengthen the idea of the verb” (G-K 
§ 113 1). This is extremely rare in pure Aramaic.’ In the LXX there 
are two main forms adopted by the translators for rendering this idiom. 
Leaving on one side the purely Greek constructions, in which an adverb 
replaces the infin., or the infinitive is dropped without replacement, 
and on the other the solitary instance of barbarously literal employment 
of the Greek infinitive (Jos 1713 B), Thackeray (Gr. 48 f.) shows that the 
translators had recourse to— 


(a) Finite verb with dat. of the cognate noun. So Gen 215 βρώσει 
φάγῃ = INA Sox, 1τ θανάτῳ ἀποθανεῖσθε -ΕΥ Τὶ NAD, 

Possible NT examples : 

1. Lk 2215 ἐπιθυμίᾳ ἐπεθύμησα. 

Ac 23° ὅρκῳ ὦμοσεν (not citation, but in introducing LXX quotation), 
417 (EP syr", Chrys.) ἀπειλῇ ἀπειλησώμεθα, 5° παραγγελίᾳ 
παρηγγείλαμεν, 2513 ἀναθέματι ἀνεθεματίσαμεν. 

Also the following in LXX citations: Mk 710 -- ᾷΝῦὺ 154, Mt 13%, Ac 217, 
With qualifying adjective, Mk 5% ἐξέστησαν εὐθὺς ἐκστάσει μεγάλῃ, Lk 15 
ACD ἀνεφώνησεν φωνῇ μεγάλῃ (κραυγῇ NBL). (With this BGU iu. 
42722, καὶ βεβαιώσει πάσῃ βεβαιώσει, has been compared. But this stereo- 

1 See Dalman WJ 34. Torrey (CDA 33) disputes the rarity. 
VOL. PART Pl ——30 


444 A GRAMMAR OF NEW TESTAMENT GREEK. 


typed formula which occurs in scores of contrasts is not parallel to any 
of these examples, but rather to Eph 15.) 


2. Jn 32° χαρᾷ χαίρει. (Jn 1883 2119 must not be placed under this 
heading.) 
4. Jas 5" προσευχῇ προσηύξατο. 


See the discussion in Prol.? 75 f. 245, and more fully in Hinl. 118 n.', 
where P Oxy i. 5!° (early Christian document iii/iv A.D.) is mentioned, 
ὅτι δοχῇ δεκτικόν ἐστιν. Radermacher (Gr.? 129) adds Anderson-Cumont 
Grégoire Studia Pontica ut. 71α. 1. βιώσασα Bio, and calls attention 
to the many parallels in the language of Attic tragedy and the Old 
Comedy, ¢.g. φόβῳ ταρβεῖν, φόβῳ δεδιέναι, φύσει πεφυκέναι, νόσῳ νοσεῖν. 
He rightly sees the Semitic influence only in the extension of such ex- 
pressions in the NT. 


(b) Finite verb with participle of the same verb or a verb of kindred 
meaning. Thus Gen 3! πληθύνων πληθυνῶ-: ΠΞ δ 137, The only 
decisive NT examples are in quotations from LXX: 


1. Mt 13'* (=Mk 4:3) βλέποντες βλέψετε (LXX). 
Ac 734 ἰδὼν εἶδον (LXX). 
3. [1 Co 2! (so Wendt on Ac 733. But the only possible meaning is 
‘““When I came”’.] Eph 5° tore γινώσκοντες. (See Prol.® 
245, also supra, 22 and 222.) 
4, Heb 6" εὐλογῶν εὐλογήσω σε καὶ πληθύνων πληθυνῶ oe (LXX). 


Canon Box (OA ii. 547) calls attention to the extreme frequency of 
this Hebraism in 4 Ezra as one of the reasons for postulating a Hebrew 
original behind the Latin text. He instances 45 excedens excessit cor 
tuum, 435 festinans festinat, 5°° odiens odisti. Another example illus- 
trates (a) above: 457 mensura mensuravit tempora et numero numeravit 
temporada. : 

See discussion in Prol. 761 and Hinl. 118. 

A good Ko.v7 instance is P Tebt ii. 4211 (iii/A.D.) καὶ μὴ σκύλῃς τὴν 
γυναῖκά σου ἢ τὰ παιδία, ἐρχόμενος δὲ ἔρχου is Θεογονίδα, “ but in any case 
came to Theogonis.” (The editors’ trans. ““ when you come,” misses 
the pleading note of the urgent appeal.) See Goodspeed AJT7’, xii (1908), 
p. 2491. With this cf. instances of redundant participles in class. Gr. 
given by K-Gii. 99. (Radermacher Gr.* 210, cft. Schol. Dem. c. Androt. 
17, ἀποδρὰς ᾧχετο, and Bekker Anecd. 425: 5 ἀπιὼν ᾧχετο᾽ συνήθης ὁ 
πλεονασμὸς τοῖς ᾿Αττικοῖς.) This may explain Eph 5°. In the LXX 
Hebraism is undeniable. 

(2) The cognate accusative of the abstract noun is sometimes used 
in the LX X, where the Hebrew uses a similar construction to strengthen 
the force of the verb, e.g. Gen 2733 ἐξέστη δὲ Ἰσαὰκ ἔκστασιν μεγάλην 
σφόδρα. The construction is common to Hebrew and Aramaic. But 


1 Against the parallel from Aeschylus given there, Mr. C. D. Chambers 
wrote (letter, Aug. 1921), “‘ The passage in P.V. would only be even remotely 
parallel if it ran. μάτην βλέποντες, μάτην ἔβλεπον, οὐ κλύοντες οὐκ ἤκουον." 


SEMITISMS IN THE NEW TESTAMENT. 445 


ἐφοβήθησαν φόβον μέγαν (see Mk 45) is only given twice in H-R (Jon 
110. followed v.'® by ἐφ. φόβῳ μεγάλῳ, 1 Mac 108) as against φοβεῖσθαι 
σφόδρα twenty times. 

Examples in NT of this intensive use, resembling the Semitic con- 
struction : 
1, Mk 41 ἐφοβήθησαν φόβον μέγαν. 

Mt 210 ἐχάρησαν χαρὰν μεγάλην σφόδρα. 
2. Rev 109 ἐκαυματίσθησαν καῦμα μέγα. 

(3) The adverbial use of an auxiliary verb (171 28, 221). 


(a) The clearest instance is the use of προστιθέναι (προστίθεσθαι) 
for the Hebraic ἢ" ΟῚ (ec. infin.)=7aduv. Thackeray (Gr. δ2 1.) shows 


that in LXX it takes three forms: (a) προσέθετο λαβεῖν, (8) προσέθετο 
καὶ ἔλαβεν, (y) προσθεὶς ἔλαβεν. (a), with 109 examples, and (8), with 
9, are direct imitations of the Hebrew; (+), with only 6 instances, is nearer 
to classical types. 

Varieties of the same Hebraism appear in the Latin text of 4 Ezra 
(see G. H. Box, in OA ii. 548), e.g. 5°? adiciam (dicere) coram te, 855 noli 
ergo adicere inquirendo, 941 adiciam dolorem, 1019 adposui adhuc loqut. 


(a) is represented in NT by— 
1. Mk 1455 Ὁ οὐ μὴ προσθῶ πεῖν (al. οὐκέτι οὐ μὴ πίω). 
Lk 204? προσέθετο πέμψαι (not D)=Mt, Mk πάλιν ἀπέστειλεν. 
Ac 12° προσέθετο συλλαβεῖν καὶ Πέτρον is rather similar, but see 
BI-D § 435, and οἵ. Clem. ad Cor. 127 καὶ προσέθεντο αὐτῇ 
δοῦναι σημεῖον. 


(γ) Lk 19" προσθεὶς εἶπε (cf. Apoc. Petr. 4 προσθεὶς ἔφη). 


See Prol.? 233, Hinl. 3, Vocab. 551. 

Also Schmidt De Eloc. Joseph. 514-7, Crénert in Wessely Studien 
iv. 3, Deissmann BS 67 n., Burney Aram. Orig. 14. 

Helbing (Gr. p. iv) disputes that (a) is necessarily Hebraistic and cft. 
Pseudo-Callisthenes m1. 41 (end), οὐκέτι οὖν προσεθέμην ἀδύνατα ἐπιχειρεῖν, 
also P Grenf i. ὅ339, (But the obscure Greek of this iv/a.p. letter 
only furnishes a parallel in Cronert’s rdg., προστεθείκαμεν ἐᾷν.) 


(6) Wellhausen offers two other examples (W!1 28, 221). 

Mk 148 προέλαβεν μυρίσαι. W.C. Allen (Comm. Mark 169) regards 
constr. as unclassical, as προλαμβάνω has nowhere else the sense of “ an- 
ticipating ’’ the action of a subsequent verb. He refers to Joseph. Ant. 
vi. 13. 7, xviii. 5.2; B.J.i. 20.1; Ignat. Eph. 3%, but thinks they are 
not parallel. He concludes that, while the phrase is not impossible 
Greek, it is probably a translation of Aram. root o1p. Mt 26% avoids 
προέλαβεν. Lagrange (S. Marc, in loc.) thinks this on a par with the 
Attic φθάνω c. inf. (rare in Att. but frequent in later writers) and cft. 
Jos. Ant. xviii. 9. 7 φθάσας ὑπαντιάζειν. But he admits the striking 
resemblance to Aramaic and compares syr‘'" N!9D2 N17), 


Mk 629 ACDw lat syr ἀκούσας αὐτοῦ πολλὰ ἐποίει he listened 
to him often. Debrunner (BI-D ὃ 414) agrees, but W. ὦ. Allen and 


446 A GRAMMAR OF NEW TESTAMENT GREEK. 


Lagrange prefer ἠπόρει (NBL boh) as giving better sense (“car avec 
ἐποίει ce qui suit est absolument banal”). Strangely, neither com- 
mentator recognises a possible Semitism. 

The two exx. under (b) are not found in the LXX, but Thackeray’s 
comment on the adverbial use of all such auxiliary verbs may apply. 
‘The classical language had used verbs like λανθάνειν and φθάνειν with 
a participle in a similar way: in the later language the participle with 
(προ)φθάνειν was replaced by an inf.: the constructions given above may 
be regarded as a sort of extension of this use” ((r. 54). 


πολλά adverbial (Wt 28). 

“The frequent πολλά as an adverb may be due to translation of the 
Aramaic ἢ». Cf. Dn © 213 νὴ Ξε πολλή, 59 615. 34-- πολύ᾽᾽ (W. C. Allen 
Exp T xiii. 330). Marti (Gr.* 92) gives Nay, pl. fem. jy much; adv. 
very, Dan. Pap. El. 1. 2 [=Strasb. 2]. Dalman (Gr.? 102) says, “΄ Peculiar 
to the Galilean dialect is a special preference for the ending in}. This 
accounts for the form }\3D very for ‘3D.” 

The NT occurrences of πολλά adverbial are given by Hawkins HS? 35 
thus : 

1. Mk 14 312 510. 23. 38. 43 620 926 153, He regards all other instances as 
accusatives. 

3. Rom 16 12, 1 Co 1612: 19. 

4. Jas 83. 

The free use of the adverbial accus. in Greek removes this from the 
category of Semitisms. See LS s.v. πολύς. The disproportionate use 
of πολλά in the second Gospel is a Marcan mannerism which may be due 
to Aramaic influence. 


Πάλιν and εὐθύς as conjunctions. 

Πάλιν, a favourite Marcan word (in narrative Mt 6, Mk 26, Lk 2), is 
sometimes an inferential conjunction, not merely again, but further, 
therewpon, and has been claimed as a translation of the Aramaic tub 
(Dalman Gr.? 218, ja;n 2ῚΠ ferner noch). So W? 28, #21, endorsed by 
Souter Lez. sv. But in many even of the Marcan instances, the 
meaning is really iterative, and where the meaning is inferential it is 
unnecessary to go back to Aramaic. ‘‘ Might not πάλιν come to have 
the same secondary meaning as ‘ again’ in English, 1.6. “ however’ ? 
See P Oxy xiv. 1676?°” (Moffatt Hapos viii. 20. 141). [Med. and MGr. 
—R. McK.] 

Εὐθύς (εὐθέως) is not only extremely frequent in Mark, but is sometimes 
an inferential conjunction (e.g. Mk 1#!: 33..59. 80 “So then”). Hawkins 
HS? 12) gives the relative occurrences in the Gospels as Mt 18, Mk 41, Lk7, 
Jn 6; in narrative Mt 12, Mk 34, Lk 1. Dalman (WJ 28) equates 
the temporal εὐθύς (-éws) with Aram. 7°, which, however, is far less 
common. Lagrange (S. Mare p. xcii) suggests that its other meaning 
resembles Ἰδὲ or j"IN3, which occurs often in Daniel (see Marti Gr.* 57*). 

Burkitt (Hv. da-Meph. ii. 89) suggests influence of Hebr. } consec. 
But Mk’s freedom from Hebraisms weakens this contention. Dalman 


SEMITISMS IN THE NEW TESTAMENT. 447 


is doubtless right in saying the excessive use of this adverb in Mark 
depends on “ the particular predilection of the author, and is due prob- 
ably to Greek rather than Jewish-Aramaic influence.” 


Temporal use of idov. 


This has been claimed as an Aramaism (W! 29, 221) representing 
Aramaic 87 (neither Marti (Gr. 67) nor Dalman (Worterb. 107) quotes 


any such use). 
1. Lk 13" ἣν ἔδησεν ὁ Σατανᾶς ἰδοὺ δέκα καὶ ὀκτὼ ἔτη. 
13? ἰδοὺ τρία ἔτη [ἀφ᾽ οὗ om. AXA... min 5.1 YS") ἔρχομαι. 

This corresponds to the Hebrew use of ΠῚ (Ges-K § 136, ὃ. 3), which 
is often translated literally by ταῦτα in LXX, eg. Gen 31% Dey bent 
yrs my LXX, ταῦτά μοι εἴκοσιν ἔτη ἐγώ εἰμι ἐν τῇ οἰκίᾳ σου. So 
Num 1452, Judg 16%, Zech 75 (7δη). But this temporal ΠῚ is sometimes 


rendered ἰδού. 

Deut 27 ἰδοὺ τεσσαράκοντα ἔτη Κύριος ὁ θεός σου μετὰ σοῦ. 

84 οἱ πόδες σου οὐκ ἐτυλώθησαν ἰδοὺ τεσσαράκοντα ἔτη. 

Gen 2755 ἐπτέρνικε γάρ με ἰδοὺ [R ἤδη A] δεύτερον τοῦτο. 

It is noteworthy that where this nominativus pendens occurs in Mk 8? 
(to which D adds εἰσίν, ἀπὸ πότε) ἤδη is the word used. 

The papyrus instance, BGU iii. 948, given in Proleg.® 11 n.1, is very 
close to Lk 1316, which, however, may well be an instance of Luke’s - 
adoption of LX X language. 


6. VERBS. 

(i) Voice. 

The Passive is used less freely in Aramaic than in Greek. Wellhausen 
looks for evidence of Aramaic influence in three directions (W! 25, 718). 
(a) Aramaic avoids passive when the subject of the action is named. 
Hence comparative rarity of passive with ὑπό in Mk. 

M-G show ὑπό after passive verb as follows : 


1. Mk 7 times, Mt 22, Lk 22, Ac 36. 

Ὁ: “ἢ 1: 59. 9π|1: Εἰον Ὁ 

3. Rom 3, 1 Co 12, 2 Co 10, Gal 4, Eph 2, Phil 2, Col 1, 1 Th 2, 2 Th 1; 
2 Timi I 

4. Heb 8, Jas 4, 1 Pet 1, 2 Pet 5, Jude 2. 


(6) Impersonal use of 3rd plur. act. in place of passive. This is 
usual in Hebrew (G-K ὃ 144 g) as well as Aramaic (Kautzsch Aram. Gir. 
§ 96. 1 (c)). 


1. Mk 614 1015 1326 1527; Μ 123 515 716 949. 
ΠΕ 05: 22028 1425 169) W725) 1.835. 295. Ac ΟΣ ΠῚ Ξ 125: may be de- 
tached from its context in Q. Cf. Mt 10'.] 


448 A GRAMMAR OF NEW TESTAMENT GREEK. 


2. Jn 15® 207, Rev 128. 
3. 1 Co 1029 BDG. [An echo of several passages in LXX.] 
4. Heb 101 (see Proleg. 58 f.). 


(c) The use of an intransitive verb in place of the normal Greek passive. 


1. Mk 431: 719 (contr. Mt 151”) 94 (contr. v.4”) 1471. 
Mt 1727 812 (contr. Lk 1378). 
Lk 45 85. 


(W. C. Allen, Exp T xiii. 330, found in ἔρχεται Mk 451 a mistranslation 
of the Aphel or Ittaphal of xn» “ bring” or“ be brought.” Lagrange, 
S. Marc p. xcvi, proposes that the text should read ΝΠ 3rd plur. Aph. 
as in Dn 5%, and that the 7 was mistaken for the interrogative particle.) 

Archdeacon Allen has strengthened the case for Mark’s Aramaism 
by giving a list of thirteen places where Mt. has changed an active or 
middle verb in Mk. into a passive (JCC, Matthew p. xxiii). 

The three groups of data given above vary in value. The statistics 
of (a) are not very convincing, especially as the free use of ὑπό after a 
pass. verb is found in reported sayings of Jesus ; (b) is more weighty, as 
this use is uncommon in Greek apart from λέγουσι, φασί. [Yet note 
that in all the exx. from the Gospels under (0) Pallis preserves the idiom. 
The passive is rarely used in MGr.—R. McK.] 


(ii) The Infinitive. 


In the LXX the influence of the prep. ὃ has given the infin. ἃ very 


wide range. Thackeray (Gir. 24) observes the great extension of the 
inf. with τοῦ, and an enlarged use of the “‘ epexegetic infinitive.” 

As examples of the former οἵ. Gen 1835, 2 Sam 1974, 1 Chr 1118. For 
the latter cf. Dt 29% καὶ οὐκ ἔδωκεν Κύριος ὁ Θεὸς ὑμῖν καρδίαν εἰδέναι 
καὶ ὀφθαλμοὺς βλέπειν καὶ ὦτα ἀκούειν. 

Moulton (Hinl. 346) denies that such infinitival constructions in the 
NT owe anything to Semitism, except in those books where actual transla- 
tion from Hebrew or Aramaic is in evidence. 


(a) τοῦ c. inf.—To the full treatment accorded to this construction 
in Proleg. 216 f. little need be added here. Hawkins (HS? 48) classifies 
all the NT uses, and adds, “ὑπὸ telic use of τοῦ with the infinitive remains 
a decidedly Lucan characteristic.” 

Radermacher (G7.? 189) recognises its moderate employment in correct 
Greek (as also by Mt and Mk), but remarks on the frequency and freedom 
of the use in Lk, Ac, and Paul as resembling the ‘“‘ Jewish-Greek ”’ of 
the LXX. For class. reff. see Stahl Synt. 675, K-G ii. 40. Thumb 
(Dial. 373) recognises ultimate Tonic influence on Attic. 

Only one NT example deserves examination as direct Semitism. 


7 ‘ > ΄ , > -~ > “ € A \ « a 
2. Rev 12? καὶ ἐγένετο πόλεμος ἐν τῷ οὐρανῷ, ὁ Μιχαὴλ καὶ of ἄγγελοι 
ae Ase ana 
αὐτοῦ τοῦ πολεμῆσαι μετὰ τοῦ δράκοντος. 


Ῥ-..--. 


SEMITISMS IN THE NEW TESTAMENT. 449 


Charles (ICC, i. 322) rejects Moulton’s explanation (Proleg. 218), and 
shows that this is a literal translation of a Hebrew construction,! already 
followed by the LXX in several passages. H.g. Hos 9% ᾿Εφράιμ τοῦ 
ἐξαγαγεῖν -- ΟΠ DDN, Ephraim must bring forth; Ps 254 καὶ ἡ 
διαθήκη αὐτοῦ τοῦ δηλῶσαι αὐτοῖς =} pytinp inns (Vulg. et testamentum 
ipsius ut manifestetur illis); 1 Chr 9° ἀδελφοὶ αὐτῶν... rod εἰσπορεύ- 
εσθαι κατὰ ἑπτὰ ἡμέρας = ODT nya Nia OMS their brethren had to 
come in every seven days. So Eccles 3! ὅσα τοῦ γίνεσθαι -- πὸ Ws. 
‘“ Thus in the Hebrew the subject before ὃ and the infin. is in the nom., 
and the Greek translators have literally reproduced this idiom in the 
LXX.” The original Hebrew is then taken to be onde POND oxo 
PMD Michael and his angels had to fight with the Dragon. (N.B.—w, 046 


omit τοῦ.) Charles claims that the same use of the infinitive (this time 
without τοῦ) accounts for the reading of A in 1310 εἴ τις ἐν μαχαίρῃ 
ἀποκτανθῆναι, αὐτὸν ἐν μαχαίρῃ ἀποκτανθῆναι, and suspects that αὐτόν is a 
corruption οὗ αὐτός. The Hebrew would be 2972 Si mar TINA WR 
naop. 

Apart from the uncertainties of the text in this passage, this is a 
useful parallel, as it enables Dr. Charles to meet an obvious objection 
to his claim that τοῦ πολεμῆσαι is a Hebraism. For τοῦ ὁ. inf. does 
not seem the most literal translation of 5 c. inf. He gives (ICC, i. 356) 


a list of the various attempts made by the LX X to reproduce this Hebrew 
idiom. Ps 329 (SB) is a close parallel to Rev 13!°(A). For the Hebrew 
constr. and a full list of examples see G-K § 114 h-k, Driver Tenses § 204. 
Guillemard, Hebraisms in the Greek Testament, p. 3, ap. Mt 25, quotes 
the Heb. and LXX of Mic 5'. “ An apt example of the practice almost 
universal, in that version, of rendering ? with infinitive, after neuter 


or passive verbs, by τοῦ with Greek infinitive; to the loss very often 
of all intelligibility or sense: e.g. 2 Sam 1974, Gen 1835, 1 Chr 1118. The 
translators appear to have concluded that a Greek idiom, which was 
the appropriate interpretation of the Hebrew idiom under certain con- 
ditions, was always to be employed as its equivalent: and so have intro- 
duced into their version renderings which are otherwise inexplicable. 
And to this we owe, in great measure, the strange and startling 
instances of the rov with infinitive, occasionally met with in the NT.” 
With the one exception of Rev 127 this sweeping assertion 
is disproved by the analysis given in Proleg. 216 ff. ‘‘ The general 
blurring of the expressions which were once appropriated to purpose ”’ ; 


1 This gerundival use of > with the infinitive is common to Aramaic and 
Syriac as well as Hebrew, and Burkitt (Hv. da-M. ii. 66) has pointed out that 
Jn 98° in syrsit is a good example of the infin. used without a finite verb to 
express “must.” The alleged Aramaised Greek of Jn is free from the in- 
fection that might be looked for, since the Syriac hadé I’metdammart bak 
is a translation of ἐν τούτῳ yap τὸ θαυμαστόν ἐστιν. 


450 A GRAMMAR OF NEW TESTAMENT GREEK. 


the tendency to substitute ἵνα c. subj. for a noun clause leading to 

the similar use of τοῦ c. inf. in a few cases; the original adnominal use 

of the genitive of the articular infinitive: these account not only for 

the examples dealt with in that analysis, but also for the six LxXxX 

passages (viz. Lk 419, Ac 1347, Rom 11°, Gal 31°, Heb 10’, 1 Pet 31°. The 

LXX is not accountable for this construction in the Pauline mosaic 

of Rom 118). 

See also Radermacher Gr.2 188 ff., Ὁ. Emrys Evans CQ xv. 26f. 
(Vide infra, pp. 484 f.) 

(b) The simple infinitive (a) in jussive sense, corresponding to late 
Hebrew independent infin. ὁ. 2, 

1. Lk 24” καὶ κηρυχθῆναι ἐπὶ τῷ ὀνόματι αὐτοῦ μετάνοιαν. Thus W? 23, 
Das Ev. Iucae, 141. The sense precludes the dependence 
of the infin. on οὕτως γέγραπται, as seen by syr*” and 
arm., which substitute ἔδει for γέγραπται, and by AC*fq 
vg., which insert καὶ οὕτως ἔδει after γέγραπται. But 
possibly the infin. depends, by zeugma, on the εἶπεν of v.*°. 
See (8) below. 

2. Rev. 131° A. See (a) above. 

(8) After εἶπεν. This is not set forth in G-K or in Driver Tenses, 
but W. C. Allen (Mark, p. 50) shows that in late Heb. and Aramaic 
TX = command, was followed by ? ὁ. infin., and is represented in LXX 
or @ by εἶπεν c. infin. Thus 1 Chr 2117, 2 Chr 118 143 2921. 27. 80 314. 11 3521, 
Bsthel= Goin 2216) 228 613!" 52: 

1. Mk 5% καὶ εἶπεν δοθῆναι αὐτῇ φαγεῖν. 

87 καὶ εὐλογήσας αὐτὰ εἶπεν καὶ ταῦτα παρατιθέναι. 

Lk 1218 εἰπὲ τῷ ἀδελφῷ μου μερίσασθαι μετ᾽ ἐμοῦ τὴν κληρονομίαν. 

Allen grants that the usage in Mt 1013 and Lk 9° is not quite parallel. 
3. Rom 2” ὁ λέγων μὴ μοιχεύειν μοιχεύεις ; 

But Lk 1213, Rom 233 (cf. Mt 5%4: 39) are sufficiently close to P Fay 
109%, cited in Vocab. 372a, to remove them from this category. For 
Mk 54 see BI-D ὃ 392 (4). We may add that the simple inf. in jussive 
sense after λέγω and εἶπον is quite classical: v. LS. 

(c) ἐν τῷ 6. infinitive—Regularly used in LXX to render 3 c. inf. 
According to Dalman, WJ 33, the Targums copy the Hebrew idiom, in 
Biblical Aramaic the kindred construction of 3 ο. inf. is used (Dan 67+), 
but the construction was wanting in spoken Aramaic. 

The NT occurrences are: 

1. Mk 44 (=Mt 134=Lk 85) 648. 


Mists 20 2 715 
Lk 18.321 96.27.43 921 51.12 δ. 40. 4. 018. 29, 33. 34.36.51 1036. 38 


111- 27, 37 1915 141 1711. 14 1835 1015 944. 15. 80. 51. 
ἌΣ πο 30, 86 03.115. 101: 
3. Rom 34 (LXX) 1518, 1 Co 1521. Gal 418. 
4, [Heb 28 8315. 156 813. But none of these clearly temporal. ] 


SEMITISMS IN THE NEW TESTAMENT. 45] 


Mk 44 and the majority of the Lucan examples are found in καὶ 
eyévero constructions (see tables p. 426 above). All the above passages 
use ἐν τῷ ὁ. inf. in the temporal sense (including Lk 1215, see Moulton 
Einl. 342 n.1) except Ac 37° 48°, Ro 158 and those from Hebrews. 

The treatment of this construction in Proleg. 14, 215 was slightly 
modified in Proleg.? 249 (Hinl. 341) under the influence of E. A. Abbott, 
who wrote (Nov. 1907): “‘ Of course ἐν τῷ c. infin. =consisting in, etc., 
would be allowable in Attic. But I confess I go with Blass in thinking 
that ἐν τῷ ο. inf.=during is non-existent or very rare in Thucydides.” 
The instinctive feeling of two such scholars as Dr. Abbott and Dr. Henry 
Jackson led to the transference of this ““ Hebraism” to the category 
of “‘ possible but unidiomatic Greek.’ This, of course, does not apply 
to Ac 3°6 48° or the examples in Hebrews, which do not depart from 
classical usage. Moffatt’s rendering of Rom 151% “ with all joy and 
peace in your faith,” confirms us in removing this also from the class of 
Semitisms. The remarkably large number of examples in Luke and 
Acts is one more evidence of the great influence of the LXX on the style 
of Luke. 

That the temporal sense of ἐν τῷ c. inf. is not impossible Greek seems 
to be shown by Soph. Ajax 554 ἐν τῷ φρονεῖν yap μηδὲν ἥδιστος Bios. 


(iii) The Participle. 


(a) The Use of the Participle in Periphrastic Tenses.—See the very 
full discussion of this question in Prol. 226f., where the periphrastic 
imperfect is recognised to be a secondary Semitism in the Synoptics and 
Acts (ce. 1-12), inasmuch as these books are based on direct translations 
from the Aramaic. Blass’s treatment (Gr. 202 ff., BI-D ὃ 353) is on the 
whole accepted (as also by Thumb Hellen. 132). The construction is 
classical enough in itself (see K-G i. 38 ff.), but with a certain emphasis, 
that justifies its use in John and Paul, but can hardly be maintained else- 
where in the NT. Moulton’s papyrus examples of ἔσομαι c. perf. part., 
and of the periphrastic pluperf. are supplemented in Hinl. 358, and more 
might be given. Here we must only stop to note Mt 2459 πεσοῦνται 
as a correction for the more vernacular ἔσονται πίπτοντες of Mk 13”. 
[Med. but not Μόν, though found in Laconian dialect.—R. McK. ] 

Restricting our survey to the periphrastic imperf. we notice that 
this is fairly frequent in the LX X (see Conybeare and Stock, Selections, 
p. 69). The construction is found in Hebrew, though its frequent use is 
a mark of the later writers, and of the decadence of the language (Driver 
Tenses, § 135 (5)). In Aramaic, however, this analytic tense often super- 
sedes the imperfect. In Biblical Aramaic the periphrastic tense rather 
emphasised the duration or the repetition of the verbal action or condition 
(Kautzsch Aram. Gr. ὃ 76 (f), Marti Gr.2 ὃ 102 (d)). Duration was 
emphasised by this use in the Aramaic of the Babylonian Talmud (Margolis 
Gr. ὃ 58 (ἢ). The same construction is rarely used in the Targums, but 
is common in the Palestinian Talmud, though generally to bring out the 
thought of duration, repetition or habit (Stevenson Aram. Gr. § 22 (2)). 


452 A GRAMMAR OF NEW TESTAMENT GREEK. 


In the NT we find the imperf. of εἰμί with the present partic. as 
follows : 


its Mk 112 96. 18 438 δ. 11 94 1032. 32("15) 143. 40. 49. 54 1549. 43 
Mt 72° 830 1922, 
Lk 110. 21, 22 933. 51 420. 31. 38. 44 51. 1θ(" 15) 29 613 810 G53 1114 1310. Al(his) 
144 15! 1977 9137 938. 53 9413. 32. 
Ac 110. 13, 14 22. ὃ. 42 Sl. 13, 28 99. 28 1051. 890 118 195. 6, 20 147 169. 12 
187 213 2219. 20, 
He Jn 19. 28 96 323 1039 11: 1328 1818. 25. 30. 
3. (2 Co 51°) Gal 122- 28) Phil 255. 
4, 1 Pet 25. 


The most important results from an analysis of these data are Mt’s 
almost complete avoidance of this locution when following Mark’s narra- 
tive, Luke’s rejection of it in every instance where a Marcan parallel 
allows comparison, and yet the freedom with which it occurs in the Lucan 
writings. It isso often introduced by Luke when absent from the Marcan 
source that one hesitates to suggest its frequency in ce. 1, 2,in “ Proto- 
Luke,” and in Ac 1-12 as evidence of fidelity to Aramaic originals. 
(Lagrange S. Luc. p. cv, observes that most of the examples in the Gospel 
accord with Greek usage.) 

Blass finds in most of the Johannine passages that ἢν “* has a certain 
independence of its own.” It is strange that though Burney devotes a 
page and a half to a list of the instances of this usage in the Aramaic of 
Daniel, it is merely to show that the LXX and Theodotion generally 
translate by the imperfect, whereas he is silent about the 10 instances 
that might possibly be adduced from the Fourth Gospel. 

A similar construction in which eyévero is joined with a present 
participle appears in Mk 1! 9-7 (to be changed by Mt every time). Archd. 
Allen (Exp T xiii. 328 £., 100, Matthew xxii.) quotes Dn 115 and La 11 
as evidence of its use in LXX or Theod. to render the same idiom in 
Biblical Aramaic. The construction only comes once in the Apocalypse, 
here with γίνομαι. 

Rev 3? yévou γρηγορῶν. 


ς 


(0) Redundant use of Participle-—For the various forms of this Semitic 
pleonasm see Dalman WJ 20 ff.; W1 17, 714; Lagrange S. Matthieu 
p. xev, 3. Mare p. Ixxxvii, S. Lue p. evi; W. C. Allen Exp 1' xiii 330 ; 
Burney Aram. Orig. 52 ff. 

It is hard to say when the participle is really pleonastic, but a Semitic 
flavour clings to the following examples : 

(a) Ἐλθὼν (ἐρχόμενος) ἀπελθών, πορευθείς, coupled with finite 
verb. [Med. and MGr.—R. McK.] 

1. χθών: ΝΠ >> 72° 12% 1105 0 Gtk oe. 

᾿Απελθών. Mt 1378 46 1830 2518. 25. 


1T include 2 Co 51° in spite of the disclaimer in Prol. 227. Of recent 
commentators Windisch agrees with Moulton, but Plummer, Bousset and 
Lietzmann treat qv . . . καταλλάσσων as a periphrastic imperfect. 


SEMITISMS IN THE NEW TESTAMENT. 453 


Πορευθείς. Lk 722 (=Mt 114) 1372 141° 15. (In 8" the Syr. versions 
have not translated πορευόμενοι.) 
[Ac 16%: 39 ἐλθόντες emphatic. ] 
3. Eph 217 ἐλθὼν εὐηγγελίσατο might possibly come into this class, but 
ἐλθών is probably significant. (See E. Haupt in Meyer’s 
Kommentar, in loc.) 
N.B.—This construction, corresponding to the redundant yon and 


ΝΞ in Hebrew, which is also Jewish-Aramaic, is absent 
from John. 


(8) ᾿Αφείς, καταλιπών (with verb of departure). 
1. Mik 488 813 1215 1450. 
Mt 1886 164 2127 2322. 
N.B.—Jewish-Aramaic rather than Hebrew. It is absent from Luke 
and John. 


(y) ᾿λναστάς or ἐγερθείς (followed by verb of motion). 
Se ikesl Son Le ΔῈ LOS : 
Mt 99 124 213 14, 20. 21 97. 19. 
Lk 139 429. 38. 39 528 1518. 20 1719 931 (24:2) 9458. 
Ac 56 827 059 1090. 28 9910. 
N.B.—This idiom is common to Hebrew and Aramaic (see Dalman 
loc. cit.). It is absent from John. 


(δ) ᾿Αποκριθεὶς εἶπεν. 

This locution is of interest apart from the element of redundancy. 
It is strictly redundant (in the sense that nothing has been said to which 
an answer is needed) only in a few places, viz. Mk 9° 1113 1235. Mt 1159 
1238 (Ὁ 1615) 174 28°. It deserves mention here because of its extreme 
frequency in the Synoptic Gospels, and its close resemblance to the 
common Hebrew idiom 78" jY, This Hebrew construction is copied 


by the LXX and the Targums and in Biblical Aramaic D8} 73) is often 
found. But Dalman goes on to say that in later Jewish Aramaic this 


formula is quite unknown. “Direct speech is introduced by the simple 
ἜΝ... The word for ‘answer’ in Galilean-Aramaic 1°38 is rarely 
used. . .. 2°N8, the word for ἡ answer’ used by Onkelos, appears to 


be as yet a learned term for ἡ making good an objection.’ Probability 
supports the view that the formula in question was unknown in genuine 
Aramaic.” This statement is the more significant in view of the follow- 
ing data. The locution is found (sing. or plur. with an occasional λέγει. 
ἔφη OF ἐρεῖ). 
1. Mk 15 times, but with a high proportion οὗ λέγει for εἶπεν. 

Mt 45. ,, 

Lk 38 __séo, 

Ae 5 τὰ (419 529 $24. 34 252): 


454 A GRAMMAR OF NEW TESTAMENT GREEK. 


2. In John the participial construction is not found once, but, as 
Burney shows, ἀπεκρίθη (-θησαν) occurs at asyndeton opening 65 times 
(and with ἀποκρίνεται once), whilst the verb with a connective particle 
opens a sentence 1] times. Mk 1259 is the only other instance of ἀπεκρίθη 
as an asyndeton opening. Burney’s conclusion is remarkable. ‘‘ It is 
difficult to resist the conclusion that ἀπεκρίϑη καὶ εἶπεν is a literal 
rendering of the Aramaic 78) MY and ἀπεκρίθησαν καὶ εἶπαν of 
PUNY ja, for which, as we have seen, they stand in Theodotion’s 
Daniel.” But in 38 instances ἀπεκρίθη (-noav) introduces the words 
spoken without further verb. 

In other words, the first three Evangelists have modelled themselves 
here on the familiar language of the LX X, whereas John in 26 passages 
follows the example of the Aramaic part of Daniel, and in the remaining 
38 passages uses the ordinary Hellenistic word, betraying his Aramaic 
cast of thought only by asyndeton. 


(ce) Ἐλάλησεν (εἶπεν) λέγων. 
1. Mk 8538 1226 Mt 2911. 2818 Lk 1458 9451. Ac 826 2631. 
2. Jn 822. 
(For the indeclinable use of λέγων (λέγοντες) = siond in Rev 4: 611. 12 
111-16 148 see Charles [CC, in loc.) 
The Hebrew siond . . . 727) is also imitated in Biblical Aramaic 


twice (Dn 653, Ezr 5"), and in the Targums, but Dalman denies its place 
in the later Jewish-Aramaic dialects. On the other hand, Lagrange 
(8. Matth. p. 1xxxix) finds it a dozen times in the Elephantine papyri 
and accepts it as commonly spoken, but not very pure Aramaic. 

For the constant introduction of oratio recta by dicens in 4 Kzra, 
esp. dixit dicens, 113? 124°, see G. H. Box in OA ii. 548. 

Other participial uses, such as καθίσας, ἑστώς, σταθείς, are less 
pleonastic than idiomatic, and belong to the category of Semitisms of 
vocabulary rather than of grammar. See Dalman WJ 22f., Proleg.® 
230, 241. To the same class belongs Mt 5%, καὶ ἀνοίξας τὸ στόμα αὐτοῦ 
εδίδασκεν αὐτοὺς λέγων, on which see Bornhiuser Die Bergpredigt, 6 ff. 

Wellhausen accounts for the wealth of pleonastic verbs in these con- 
structions by the lack of compound verbs in Semitic languages, which 
necessitated the use of a special root to express every shade of meaning. 


(Ὁ "Ap&apevos. [See Proleg.? 182, 240 (Hinl. 287), Vocab. 82 6.) 
1. Mt 208. 

Lk 235 2447, 

Ae 142s: 

2. [Jn] 8°. 

Torrey’s claim (CDA 25) that this is an Aramaic idiom in Acts is 
disputed by J. W. Hunkin (J7'S xxv. 401) on the ground of the papyri 
usage, also of its occurrence in Xenophon and Plutarch. See further 
np&a(v)ro 6. infin., below under iv (a), 


SEMITISMS IN THE NEW TESTAMENT. 455 


(iv) Indicative Mood. 
(a) Redundant Auxiliary Verb. 


This characteristic of Hebrew and Aramaic has already been dealt 
with under the Participle, and we there saw that in 26 instances 
John substitutes the indicative for the participle so familiar to us in the 
Synoptic formula ἀποκριθεὶς εἶπεν. The Johannine ἀπεκρίθη καὶ εἶπεν 
is not the only example οὗ a redundant verb in the indicative. The 
most striking example is ἄρχομαι. 


1. "HpEa(v)ro ὁ. infin. 
Mk 145 923 41 517.20 68. 7. 84. 56 911. 81.32 1028. 82. 41.47 1]16 1921 135 
1419. 33. 65. 69. 71 158. 18. 
Mt 41” 117-20 121 1651. 22 2632. 37.74, [Also in other tenses, in sayings 
of Jesus, 184 and 244°, and in 1439] 
Lk 421 521 715. 24. 38. 49 g12 1129. 53 191 1418. 30 1514. 24 1937. 45 999 9923 
23%, [Also in other tenses, in five passages from sayings of 
John the Baptist or of Jesus, 3° 12% 1335. 26 149] 
Ac 11 24 1876 242 2785, 
2. Jn 13°. 


This cannot be called a Hebraism, for though it is found fairly often 
in the LXX a glance at H-R shows that it has no fixed Hebrew original. 
Sometimes it represents a word with a definite meaning (e.g. Hiph. of 
DN’), often the Hiph. of 2°n, sometimes (e.g. Gen 2%) it is without 
warrant in the Hebrew, and occurs quite freely in books without a 
Hebrew source. 

On the other hand, it is claimed that its use in Mark is due to the 
use of Ww in Aramaic as an auxiliary verb. See W. C. Allen, Comm. 
Mark, 49f., who points out (a) all the 26 instances in Mk are in narrative, 
and not one has special emphasis. (ὁ) Mt omits all but 6 of Mk’s instances, 
probably from the perception that the word was Aramaic rather than 
Greek. (c) Luke’s use is remarkable. He retains only 2 of Mk’s cases. 
Of the remaining 25, 12 are in sayings, 13 occur in narrative, 5 of which 
occur in passages with Marcan parallels. His conclusion is that the 
frequency in Mk is due to translation from Aramaic, in Luke partly to 
the Aramaised Greek of his sources, partly to a feeling that (especially 
in “ἡ began to say ᾽᾽) such phrases were quite natural in Greek. I venture 
to add that its comparative frequency in the LXX may have inclined 
Luke to its use, as was possibly the case with ἀποκριθεὶς εἶπεν (see above, 
pp. 453 f.). 

G. H. Box (OA ii. 548) accounts for the very frequent use of tncipere 
c. infin. in 4 Ezra as a literal rendering of Heb. >\Nin, and cft. the 
similar use of ἄρχεσθαι in the Gospels. 

Radermacher! calls attention to a parallel to ἤρξατο λέγειν in the 
vulgar Latin coepit dicere, c(o)eptmus ascendere, ubt coepertt lucescere. 


1 Tdg. F. xxxi. Anz. 6 (his valuable review of Moulton’s Hinleitung). The 
point is not mentioned in the recent 2nd ed. of his Neuwdéestamentliche Grammattk. 


456 A GRAMMAR OF NEW TESTAMENT GREEK. 


The fullest treatment that this idiom has received is by J. W. Hunkin 
JTS xxv. 390-402, * Pleonastic” ἄρχομαι in the New Testament. Here 
Dalman’s examples of the Aram. ὙΦ and_post-Biblical Hebrew 


nnn are examined closely, the former being found to carry the 


ordinary meaning, whilst the latter is sometimes semi-pleonastic. Test. 
Levi (Charles’s ed. pp. 249 f.) is quoted in Aramaic and Greek in support 
and Enoch Ixxxix. 42-49 (which is based on an Aramaic original) is 
shown to favour this locution strongly. Archd. Hunkin then adduces 
numerous parallels from Xenophon and two from Aristophanes. This 
leads him to the conclusion: “‘ The above parallels are quite sufficient 
to show that the usage of ἄρχομαι with the infin., as we have found it 
in Mt or Lk, is no indication that either St Matthew or St Luke was 
acquainted with Aramaic. Nor does it necessarily suggest that any of 
the documents used by St Matthew or St Luke was originally written 
in that language... .’ The loose and somewhat pleonastic way in 
which “‘ begin” is used by Mark, and in some of the Lucan records of the 
speeches of Jesus, is admitted to be due to Aramaic. 


(0) Tense. 
(a) Historic Present and Imperfect as renderings of Aramaic participle. 
a. Historic Present.—The proportionately high frequency of this in 
Mark has been claimed as an Aramaism by W. ©. Allen, and in John by 
Burney. {[MGr.—R. McK.] 
The statistics (given by Hawkins HS? 144 ff., Burney Aram. Orig. 87) 
are as follows : 


1. Mk 151 (of which 72 are words signifying speaking, 6.6. λέγει, φησίν). 


Mt 93 4“. 68 33 39 39 
kg Ἢ 6 τ: τὰ x 
‘Acris + 11 


[Ν.8.---Οὐ Mt’s 93 examples, 21 are retained from Mk, and 15 occur 
in Parables. 
Of Lk’s 9 examples, 5 come in Parables. | 
2. Jn 164 (of which 121 are words signifying speaking, e.g. λέγει, φησίν). 


It is evident that both Mt and Lk regarded this as a vulgarism to 
be removed when possible (see Proleg. 121), but Thackeray, Schweich 
Lectures, 20 ff., has shown strong reason for denying that it is due to 
Aramaism. Taking the first three books of each of the four leading 
historians he finds the historic present in Herod. 206 times, Thuc. 218, 
Xen. 61, Polyb. 40. Thus in the classical age it was common to the literary 
style and to vernacular, whereas in Hellenistic it was increasingly regarded 
as vernacular. Hawkins (HS? 213) shows that out of 337 instances in 
the LXX, 232 occur in the four books of Kingdoms, of which 151 are 
found in 1 Kgd. Thackeray’s thorough examination shows that in Mk 
as in 1 Kgd the historic present tends to come at or near the beginning 
of a paragraph. The exceptions are specially dramatic, as Mk 15*4- 37 
(pictorial). Verbs of (a) saying, (b) seeing, (c) coming and going, (d) bring- 


SEMITISMS IN THE NEW TESTAMENT. 457 


ing and sending, are conspicuous. ‘The tense as a rule is dramatic in the 
sense that it serves to introduce new scenes in the drama. Cf. stage 
directions, “‘ Enter ...,” “Scene...” “ Loguitur....”? He con- 
siders Archd. Allen’s claim for Aramaism in the Marcan use quite 
untenable. ‘‘ Would he maintain that 1 Samuel lay before the Greek 
translator in Aramaic ?” 

Allen pointed out (following Néldeke, Syr. Gr. 190)! that in Syriac 
this participial expression of action described as taking place is practic- 
ally limited to the verb “‘ to say” (Eup T xiii. 329). 

8. Imperfect.—Stevenson Aram. Gr. 56. In Palestinian Talmud and 
Midrash ἢ the use of participles in place of perfect tenses in narratives of 
past events is very characteristic.” “Τὴ Old Test. Aramaic the par- 
ticiple . . . is a very frequent alternative to a perfect in narratives of 
past events. It is also used as a progressive tense descriptive of events 
in the present or the past, and in stating general truths.” In the Tar- 
gums participles often represent Hebr. imperfects, but not to the same 
extent as in the Palestine Talmud and not in the same uses. 

Allen (ué supr. also ICU, Matthew, p. xxiii) observes that there are about 
220 imperfects in Mark, about 30 of which Mt changes to aorist. He finds 
a striking parallel in Theodotion’s version of Daniel (which contains 149 
verses from Hebr. and 206 from Aram.). “* In rendering Hebr., Theod. 
uses about 9 imperfects; in rendering Aram., about 64. Of these 64 
about 4 correspond to perf., about 12 to imperf., but about 27 to a part., and 
about 21 to a part. with mn, That is to say, a literal translator, where 
he had an Aram. partic., or partic. with 11, thought it natural to 
render them by imperfects.” 

To estimate the force of this argument we must examine the other 


historical books in the NT. Hawkins’s figures for the imperfect (ex- 
cluding ἔφη, and with Burney’s correction of that for John) are: 


Mt 79, Mk 222, Lk252, Ac314, Jn 165, 


and he adds the explanation that the smailer proportion in Mt and Jn is 
partly due to the larger amount of discourse in proportion to narrative 
which they contain (HS? 51). 

If we adapt these figures to an average per WH page, the results are: 


Mt 1:16, Mk5-4, Lk3-5, Ac45, Jn3-l.? 


A further test reduces the significance of this argument for Aramaic. 
“1 find that in Milligan’s Selections from the Greek Papyri there are 22 
impf. to 111 aor.; in Mk 1, 19:39; in Mt 3 and 4, 7: 29, in Polybius 
(7 pages in Wilamowitz Lesebuch) 37:54; and in Appian (6 pp. in do.) 
90:25. So Appian here uses impf. seven times as much as Mk does, 
and Polybius 1} times.” (Note by J. H. M.) 


1 Noldeke Syr. Gr.2 206: ‘‘ Die Erzithlung verwendet das Part. act. (als 
Praes. histor.) fast nur bei ’dmar, aber dies ’dmar, ’dm*rd, ’dm*rin, ’dméran, 
‘er, sie sagte’; ‘sie sagten’ ist sehr hiufig.’’ 

2 Burney counts 118 cases in Jn 4-12, 1.5. 4°37 per WH page. 


458 A GRAMMAR OF NEW TESTAMENT GREEK. 


(b) Future for Imperative. 


Lagrange (S. Matthieu, p. xcv) traces this use in the Gospels to the 
Semitic use of the imperfect for both jussive and future, calling attention 
to the alternation of imperatives and jussives in Dan 411-18, 

Thus he accounts for Mt 54° ἔσεσθε (Lk 6386 γίνεσθε), 6° οὐκ ἔσεσθε, 
2026. 27 ἔσται (so Mk 10%: 44, but Lk 2276 γινέσθω), 23" ἔσται (so Mk 985). 

On the other hand, Mt 218 (=Lk 1931) ἐρεῖτε, where Mk 11° has imper. 


εἴπατε. 
(ec) Aorist for Present. 


In Proleg. 134. reason was given for regarding a number of NT 
examples, which seem to come under this heading, as either epistolary or 
gnomic aorists, or else as instances of the very old use (ordinary in early 
Sanskrit) of the aorist of the proximate past. 

Two instances stand apart from the others. 

Mk 1" (=Mt 3!7=Lk 322) ἐν σοὶ εὐδόκησα. This may be explained as 
summary aorist (referring to the “‘ 30 blameless years,” so G. G. Findlay), 
or the aor. of indefinite time reference, or else as the aor. of proximate 
past (the Dove the sign). All these are fully in accord with Greek usage. 
Notice, however, the allusion to Is 421 ‘WB) NNYI Wa, LXX, ὁ ἐκλεκτός 
μου, ὃν εὐδόκησεν ἡ ψυχή μου (B προσεδέξατο αὐτὸν ἡ Ψ. pov.), quoted 
again (with ὁ ἀγαπητός for ὁ ἐκλεκτός) in Mt 1218, Here the Hebr. perf. 


(delighteth, NS) is stative) is represented by Greek aor. Allen (JCC, 


Matthew 29) says, “‘ The aor. εὐδόκησα is modelled on the aorists of the 
LXX in this passage, which were probably interpreted as implying the 
divine election of Israel, and so here the divine election of the Messiah.” 

Mt 232 ἐπὶ τῆς Maicéws καθέδρας ἐκάθισαν οἱ γραμματεῖς καὶ οἱ Φ. One 
naturally remembers Ps 1! with the three perfects Jon Woy aw rendered 


in the LXX by three aorists ἐπορεύθη, ἔστη, ἐκάθισεν ; cf. Ps 122°. For 
the whole subject see Driver Tenses §§ 12, 35. 

Prof. G. C. Richards, reviewing Proleg. in JTS x. 284, wrote: “ Is 
it ingressive, ‘ they came to sit,’ effective, ‘they seated themselves,’ or 
constative, ‘they sat’ ? It would be very inappropriate to call it gnomic. 
Wellhausen claims it as an Aramaism. Is it not probable that he is 
right, and that a limited number of verbal forms in the NT will defy 
analysis on Greek lines?’ To which Moulton replied (Hinl. 220 n.*): 
‘“* T am not sure of this, although, as often shown, I am in no way opposed 
to the assumption of ‘ translation-Greek.’ It may be translated by ‘ they 
seated themselves,’ an act of indefinite time reference in the past, as every 
rabbi in turn claimed this ea-cathedra authority on beginning his life-work ; 
the tense is then natural. It is ingressive, and expresses the self-asser- 
tion of the would-be Moses more vigorously than the present could ; and 
it is iterative, for it applies to many individual scribes.” 

Wellhausen! admits that ἐβάπτισα Mk 18 (--βαπτίζω Mt 314, Lk 31) 
and ἐδίστασας Mt 1433: are rather different, as they refer to an action 


1 See W?1 25, 218. 


SEMITISMS IN THE NEW TESTAMENT. 459 


completed in a moment, or, as we should prefer to describe the tense, 
“the aorist of the thing just happened ”’ (proximate past). 


7. Nouns. 


The peculiar idioms in the treatment of nouns come more conveniently 
under other headings. One or two special uses are mentioned below. 

(1) Casus Pendens: see above, A 3. 

(2) Accusative Case: see under adverbial expressions, B5. A special 
instance of alleged Hebraism is Mt 415 ὁδὸν θαλάσσης for D' ὩΠ, 


But see W. C. Allen (JCC in loc.), who shows that it is not from LXX, but 
probably due to careless copying from some other version. 

(3) Genitive Case: see under adjectival substitutes, B 3. 

A Hebraic use of the genitive after a perf. pass. part., or a verbal adj., 
has been suspected in such passages as : 


1. Mt 25% οἱ εὐλογημένοι τοῦ πατρός (MN FA, LXX εὐλογητὸς κυρίου, 
ΟΥ εὐλογημένος ὑπὸ κυρίου). 

Lk 251] τὸ εἰθισμένον τοῦ νόμου. [Βυῦ 15 quite class. Ξετὸ ἔθος τοῦ νόμου. 

Mt 11" (=Lk 728) ἐν γεννητοῖς γυναικῶν (οὗ. Job 141 1514 254, γεννητὸς 

γυναικός, TWN 335) and contr. Gal 44 γενόμενον ἐκ γυναικός). 


2. Jn 6% (LXX) καὶ ἔσονται πάντες διδακτοὶ Θεοῦ (=Is 544 MN 7319). 


3. 1 Co 2138 οὐκ ἐν διδακτοῖς ἀνθρωπίνης σοφίας λόγοις, ἀλλ᾽ ἐν διδακτοῖς 
πνεύματος. 


Τῇ in some of these we must acknowledge the influence of the LX X, we 
can also see an extension of a use common in poetry, e.g. Soph. Ajax 807 
φωτὸς nratnuévn, Eur. Or. 497 πληγεὶς θυγατρός. See K-G i. 376.1 

(4) Dative Case: see under adverbial substitutes, B 5. Α few special 
uses remain for treatment. 


(a) Dat. of agent after pass. part. is claimed by Wellhausen as 
Aramaic, ‘“‘as amandus mihi, or rather amatus mihi” (171 25, 718), 
Noldeke (Syr. Gr.2 ὃ 279) attests the fondness of Syriac for this construc- 
tion. The one NT example is Lk 23" ἐστὶν πεπραγμένον αὐτῷ. It is 
not easy to see why Lk should have preserved the Roman procurator’s 
statement in Aramaised Greek. But the numerous examples of the dat. 
of the agent after a perfect passive given by K-G i. 422 show that it 
was perfectly good Greek. Cf. Thuc. i. 51,118; Hdt. vi. 123; Isoc. iv. 4; 
Lys. xxiv.4; Xen. Anab. τ. viii. 12, vi. vi. 32; Cyr. vu. il. 15, and more 
closely Dem. c. Aphob. 1: δεῖ διηγήσασθαι τὰ τούτῳ πεπραγμένα περὶ ἡμῶν. 
Cf. Kalker 279: “‘ Dativus, qui cum passivo conjungitur, eadem notione 
qua ὑπό c. genit., saepissime a Polybio usurpatur: e.g. i. 13. 3.” 
P Petr ii. 13(19)!* (iii/B.c.) (=Witk. 813) οὐθέν σοι μὴ γενηθῆι λυπηρόν, ἀλλὰ 


“᾿ x , m*4 , ~ , 9, 
πᾶν €[poi ἔστ]αι πεφροντισμένον τοῦ σε γενέσθαι ἄλυπον. 


1 With Jn 645, 1 Co 213, Mr. Εἰ. E. Genner cft. Soph. Hl. 343-4 νουθετήματα 
κείνης διδακτά, and such phrases as τὰ πεπολιτευμένα αὐτῶν in the orators. 


VOL. II. PART III.— 31 


460 A GRAMMAR OF NEW TESTAMENT GREEK. 


(Ὁ) Dat. after ¢voyos,.claimed by Wellhausen as “‘ ungriechisch.’’ 
(W! 33 ἢ, omitted in ed.?). Mt 5%? ἔνοχος ἔσται τῇ κρίσει. 

J.C. James, Language of Palestine 72 n.4, observes : “‘ It is equivalent to 
Sypns NPOPMN Onk. Gen. 261, LXX θανάτῳ ἔνοχος ἔσται. But θανάτου 
has the best MS support, and it is not easy to see what bearing the Hebrew 
original N11) Ni12, or the Aramaic of Onkelos, has upon the alleged Sem- 
itism of the Greek idiom. Moulton has shown conclusively (Hinl. 373 n.1 
Vocab. 217) that this, as well as other NT uses of ἔνοχος, accords with 
Greek usage. 


(c) Dr. Charles explains the difficult dative in Rev 218 τοῖς δὲ δειλοῖς 
καὶ ἀπιστοῖς . . . TO μέρος αὐτῶν as a reproduction of the Hebrew idiom 
by which 5 introduces a new subject (JCC i. p. exlviii, ii. 216 η.1)- 
Viteau Etude ii. 41f., cft. 2 Es 1012 καὶ πᾶσι τοῖς ἐν πόλεσιν ἡμῶν. .. 
ἐλθέτωσαν. 


8. PREPOSITIONS AND IMPROPER PREPOSITIONS. 


The extensive use of many prepositional phrases in LXX is recognised 
by Thackeray as due to Hebrew influence. 


᾿Από and ἐκ in constructions influenced by 12, 


(a) Thackeray (Gr. 46) notes that in LXX azo c. gen. is thus used 
instead of acc. after αἰσχύνεσθαι, εὐλαβεῖσθαι, λανθάνειν, προσέχειν, 
τρέμειν, ὑπερηφανεύεσθαι, ὑπερορᾶν, φοβεῖσθαι, φυλάσσεσθαι. 


NT instances are :- 


1. Mt 715 1017 166: 11. 12 (—Lk 121), Lk 2046 προσέχειν ἀπό. 
Mt 1028=Lk 12! μὴ φοβηθῆτε ἀπὸ τῶν... . Hawkins (HS? 64) observes 
that against this NT example the LX X contains the idiom 49 times, 
25 of them with ἀπὸ προσώπου, out of about 440 occurrences of the 
verb. 
Mk 8158 1238 βλέπειν ἀπό (see p. 31 above). 
Lk 1215 φυλάσσεσθε ἀπό. (Cf. Xen. Cyr. τι. iii. 9, Hell. vir. ii. 10.) 


2. 1 Jn 5% φυλάξατε ἑαυτὰ ἀπὸ τ. εἰδ. 


28 ‘ > - tes > ~ > - , > a 
2 μὴ αἰσχυν θῶμεν απ αὐτου εν Τῇ παρούσιᾳ αὐτου. 


W? 32 attributes φεύγειν and φοβεῖσθαι ἀπό to Semitic influence, but 
W? 25 concedes that they can also be Greek. So κρύπτω ἀπό is as 
old as Homer (Od. xxiii. 110) (BI-D § 155. 3). For ἀπὸ προσώπου, 
however, see below. Mr. Emrys Evans touches upon this idiom in a 
paper upon Case-Usage in the Greek of Asia Minor (CQ xv. 28). He cites 
C. and B. ii. p. 565 (no. 466), ἐὰν δέ τις μὴ φοβηθῇ τούτων τῶν καταρῶν. By 
the side of this ablatival genitive he sets Pelagia x. 12 (Usener, p. 12), μὴ 
δειλιάσῃς ἀπ᾽ αὐτοῦ, and concludes that the NT construction is a develop- 
ment quite natural to Greek—a “ἡ secondary”’ Hebraism. [Προσέχω and 
φυλάσσω ἀπό in Med. and MGr, βλέπω ἀπό in Med. Gr, φοβοῦμαι ἀπό in 
MGr, which prefers φοβ. c. acc.—R. McK.] 


SEMITISMS IN THE NEW TESTAMENT. 461 


(6) Mt 11°=Lk 7° δικαιοῦσθαι ἀπό is taken as Aram, DIP jd 
=Hebr. *J5'> by Wellhausen (W? 32, °25). But it seems far better to 


take this ἀπό (as so often in the Κοινή) 885 Ξε ὑπό, marking the agent after 
a passive verb. 


(c) The causal use of ἀπό has been explained thus. (For ἀπόϊη LXX 
=causal 12, cf. Gen 911, Ps 767.) W? 32, #25 cites from Synoptics : 
1. Mk 24D ἀπὸ τοῦ ὄχλου (for διὰ τὸν ὄχ.). 
Mt 187 οὐαὶ ἀπὸ τῶν σκανδάλων, 1435 284 ἀπὸ τοῦ φόβου [add 1344 ἀπὸ 
τῆς χαρᾶς αὐτοῦ]. 
Lk 2245 ἀπὸ τῆς λύπης, 2454 ἀπὸ χαρᾶς [add 1938 ἀπὸ τοῦ ὄχλου, 2176 
ἀπὸ φόβου]. 
Ac 119 ἀπὸ τῆς θλίψεως, 1214 ἀπὸ τῆς χαρᾶς, 221} ἀπὸ τῆς δόξης. 
2. Jn 218 ἀπὸ τ. πλήθους τ. ἰχθύων. 
4. Heb δ ἀπὸ τ. εὐλαβείας. 


This usage is classical (see LS). For vernacular examples see 
Kuhring 35 f., Vocab. 58f. For parallels to Mt 18’ see Prol.3 246. [Med. 
and MGr.—R. McK.] 

(d) A special kind of ablative use appears in— 

1. Mt 2774 ἀθῷος ἀπό, Ac 206 καθαρὸς ἀπό [cf. Gen 244! ἀθῷος ἀπό (€x)= 
{9 ‘pal, for which abundant parallels from papyri are given 
by Kuhring 52f. See also Vogeser Spr. d. gr. Healigenle- 
genden 26. 


Lk 24°! ἄφαντος ἐγένετο ἀπ᾽ αὐτῶν, claimed as a Hebraism by Psichari 
Essai pp. 204 ff. See Vocab. 95b. 


(6) The partitive use of ἀπό after ἐσθίειν. 
Mk 728=Mt 1527. McNeile Comm. in loc. ‘‘ The Hebraic ἐσθίειν 
ἀπό (3 53x), frequent in the LXX, is not found elsewhere in NT; cf. 


ἐσθ. ex.” It is an instance of the more general use of ἀπό or ἐκ ὁ. gen. to 
replace the partitive gen. in later Greek. Cf. MGr. δειπνάω ἀπὸ χῶμα, 
1 eat (of) earth (Thumb Handb. 102). [Very common in MGr.—R. McK.] 


(f) The phrase ἀπὸ μιᾶς Lk 1418, has been claimed as a literal trans- 
lation of Aramaic N1N 2 “ all at once,” “‘ suddenly.” See above, p. 28. 


Moulton asks (Hinl. 15 n.'), “‘ But why μιᾶς fem. ? Simply because it is 
a Greek idiom.” Blass (Gr. 140f., Bl-D ὃ 241, n. 6) regards it as a 
stereotyped phrase, ‘‘ with one mind or voice,” and cft. Aristoph. Lysistr. 
1000, ἀπὸ μιᾶς ὑσπλαγίδος (strictly of runners in a race, who rush off 
together at the fall of the single rope). Plummer [CC in loc. supplies 
γνώμης for an “ expression unique in Gr. literature,” and cft. Philo (De 
Spec. Legg. ii. p. 311), ἀπὸ μιᾶς καὶ τῆς αὐτῆς γνώμης. For the Semitic 
use of the fem. see G-K, § 122 g. I cannot trace the Aramaic idiom in 
the grammars of Marti, Kautzsch, or Dalman, or in Dalman’s W orter- 
buch. But the Syriac mech*da supplies that meaning.t Moulton (op. cit.) 


1 As Mr. C. R. North informs me. 


462 A GRAMMAR OF NEW TESTAMENT GREEK. 


adds: ‘‘ My thesis does not in the least deny the Aramaic origin: I only 
protest that the translation is quite idiomatic.” It appears that such 
idioms are not uncommon in Med. Gr. and in MGr. For one example 
(in a story from Epirus) see Thumb Handb. 240, τὸ βράδυ ἐχτύπησαν 
τὸ κούτσουρο ὅλοι ἀπὸ μνιὰ καὶ TO ἔκαναν κομμάτια, also p. 244. [Med. and 


MGr.—R. McK.] 


Διά. 


Mt 11? πέμπειν διά-- ἼὋΞ Προ Hebr. and Aram. So 171 31. The 


idiom is seen more clearly in Mk 6? διὰ τῶν χειρῶν, Ac 273 ἔκδοτον διὰ 
χειρὸς ἀνόμων, where Torrey (CDA 6) observes that the same Aramaic 
words are rendered παραδίδοται eis τὰς χεῖρας τῶν ἁμαρτωλῶν in Mk 144. 
See, however, supra, p. 29, Vocab. 145 f., where it is said that διὰ χειρός 
6. gen. is based on Ἴ3, but is not a literal translation. ““ It is obviously 
modelled upon the vernacular phrase διὰ χειρός, of money paid 
‘ directly.’ ”’ 


Eis. 


(a) Eis c. ace. in place of predicative nom. (esp. with εἶναι, γίνεσθαι, 
λογίζεσθαι) or predicative acc. (“Semitic influence unmistakable,’’ Bl-D, 
§ 157. See Oxf. Heb. Lex. p. 512). 

Moulton (Proleg. 71 1.) discounts Hebraism, “‘ for the vernacular 
shows a similar extension of the old use of «is expressing destination,” 
but he allows Semitic influence in some of the passages, adding examples 
(p. 76) from non-Semitic sources. For further examples see Radermacher 
Gr.? 21 and D. Emrys Evans CQ. xv. 241. [Med. Gr.—R. McK.] 

Johannessohn (Kasus und Prdp. 4) gives examples of this LXX 
rendering of 5. and others where LXX thus renders Hebr. nominative. 


But his most significant results show that whereas the predic. nom. with 
εἰμί and γίνομαι is common in all four books of Mac., εἰς ὁ. acc. is used 
instead of the predic. nom. in 1 Mac only, where it occurs 15 times, and 
instead of the predic. acc. 5 times. 

See F. Schulthess (ΖΝ ΤῊ xxi. 221) for denial that ᾿ predicative is 


Aramaic. Where it is found in Bibl. Aram. it is ? resultant and is a 
Hebraism. So in Syriac it is restricted to OT translation. 


1. With γίνεσθαι. Mt 214? (LXX), Lk 1319, Ac δ85, 
With εἶναι. Mt 19° (LXX), Lk 3° (LXX). 
With λογισθῆναι. Ac 1957. 
For predicative acc. Mt 2115. Ac 772 (LXX) 134 (LXX) 137? (cf. 
1 Καὶ 13" εἰς apyovra= Ἴ220). 


2. With γίνεσθαι. ὅπ 1639, Rev 8" 1619. 
With εἶναι. 1 9π 58. (But A. E. Brooke tn loc. takes it=are for the 
one thing, tend in the same direction, exist for the same object.) 


SEMITISMS IN THE NEW TESTAMENT. 463 


3. With εἶναι. 2 Co 618 (LXX), Eph 5%! (LXX). Cf. Ro 518 (εἰς xard- 
κριμα 80. ( ἐγένετο). 

With λογισθῆναι Rom 276 48 (LXX) 98, 

4, With εἶναι Heb 15 (LXX) 810 (LXX), 

(Ὁ) πιστεύειν εἰς c. acc. See Burney, Aram. Orig. 34, and Proleg. 
68, where a table is given showing the distribution of the various con- 
structions with πιστεύειν. Semitic influence is recognised in the literal 
translation of 2 ji} by the prepositional phrase, reserving the simple 
dative for δ tak 


1. Mt 188 (=Mk 942 ABLO), Ac 104 1423 194, 
2A. Jn 112 oll. 23 316. 18, 36 489 629. 35. 40 75. 31. 38, 39. 48 830 935. 36 1042 1125. 
26. 45. 48 1911. 36, 37, 42, 44. 46 141. 12 169 1729, 


1 Jn εἴν 13, 
8. Rom 1014, Gal 216, Phil 12°. 
4. 1 Pet 18, 


(c) A curious use of εἰς appears in Ac 788 εἰς διαταγὰς ἀγγέλων, where 
Torrey (CDA 33) explains that cis = δ which sometimes means “‘ according 
tox = bys PONDD ‘TPB? “by the ordering of angels.” Cf. Ps 119% 
ΤΌΞΟ» “ἢ according to thy ordinances.” 

(4) The phrase πορεύου (ὕπαγε) εἰς εἰρήνην (Mk 5%4, Lk 750 848) must 
be due to the LXX, where it often represents the Hebrew Dinvip 3p. 
See OHL, p. 516, for similar examples of this adverbial use of > with 
abstract nouns. 

(6) For εἰς ἀπάντησιν =NN rp? see Proleg. 14 n.4, [Med. Gr.—R. McK.] 


Ἔν. 


(a) For the instrumental use of ἐν see Proleg. 12, 61, 104, and above 
p-. 23; also Kuhring 43 f., Rossberg 28. [Med. Gr.—R. McK.] 

(Ὁ) The causal use of ἐν has some support from the papyri: see 
Kuhring 43, Rossberg 29, Vocab. 210. H. A. A. Kennedy (Exp 7 xxviii. 
323), however, gives a list of passages in which the LXX rendering of 2 
has probably left its influence, with the meaning because of, by reason of, 
for the sake of. 


1. Mt 6’, Ac 729 (LXX). Add Ac 2416, 

2. Jn 1689, 

3. Rom 131. 24 53, 1 Co 48 714, 2 Co 125- 9, Ph 145, Add Rom 97 (LXX). 
4, Add Heb 101° 1118 (LXX). 


(c) Dr. Moulton acknowledged a Semitism in ὁμολογεῖν ἐν (Prol. 104), 
and observes Nestle’s warning that the construction with 3 is Aramaic 


rather than Hebrew (Hinl. 169). Lagrange 5. Matthieu, p. civ, ““ En 
aram. x7 ἃ Pa. ou plutdt Aph. Le syr sin met le 3 méme ἃ Jo 12%.” 


464 A GRAMMAR OF NEW TESTAMENT GREEK. 


Moulton also accepted G. G. Findlay’s suggestion that ὀμνύναι ἐν 
and eis (LXX, but not in Pent.) are Hebraistic, occurring in religious 
formulae, and only in Mt and Rev. 


4. Mt 535. 36 9316(bis) 18 (bis) 20(bis) 21.015) 22 (bis) | 
2. Rev 108 (LXX). 


Lagrange (1.6. p. civ) ““ En gree on emploie l’accus. ou κατά, ou le dat. 
La construction avec ἐν est done sirement sémitique; en aram. yy 
a PItpe. et 3.” 

The πιστεύειν ἐν of Mk 115 Moulton afterwards accepted as transla- 
tion Greek (contra, Proleg. 67) under the influence of Dr. Burkitt, who 
compared Jer 12° μὴ πιστεύσῃς ἐν αὐτοῖς ὅτι λαλοῦσιν πρὸς σὲ καλά." 

“Then εὐδοκεῖν ἔν tive (once in Polyb.) would be encouraged by 
the same tendency, in place of regular simple dative (see Milligan Thess. 
p- 106)—the acc. also is used”’ (J. H. M.). 

To the same class belongs σκανδαλίζξεσθαι ἐν. 


1: Me i1® 1357 26%. 330 Mik 63, Lk 7. 


Lagrange S. Matthieu, p. eviii, ‘‘ Le prép. ev est un reflet du sémitique 
... L’araméen est probablement 5pn a Πέρα. Les verss. syrr. ont 
ἘΠΕῚ Sw5 comme I’héb.” 
In addition to these verbal constructions with ἐν, W! 32 224 claims 
ἔρχεσθαι év=“ to come with,” “ to bring.” 


1. Mt 21 ἦλθεν. . . ἐν ὁδῷ δικαιοσύνης Ξε γοιρσίξ the method of 
righteousness. 
1038 ἐρχόμενον ἐν τῇ Bao. a’rov=bringing his kingdom. 
Lk 23% ὅταν ἔλθης ἐν τῇ Bac. cov (BL eis=ev acc. to Wellh. Note that 
D reads ἐν τ. ἡμέρᾳ τ. ἐλεύσεώς cov). 


This interpretation seems gratuitous in every instance. 

Other uses of this preposition mentioned by Wellh. may be simply 
instances of the intrusive ἐν so common in the Kowy (Kuhring 12), 
e.g. ἐν ποίᾳ ἐξουσίᾳ (Mk 1138), ἐν δυνάμει (Mk 91), ἐν δόλῳ (Mk 14°). 
Add to these Mt 7? ἐν ᾧ μέτρῳ, 15 ἐν τοῖς ποσὶν αὐτῶν, 2215 ἐν ἀληθείᾳ 
(contr. Mk 1214=Lk 2051: ἐπ᾽ ἀληθείας. 

1. Lk 1655 ἐν πᾶσι τούτοις (NBL) is corrected to ἐπί in AD; cf. σὺν 
πᾶσιν τούτοις (2471). Possibly both represent 3, 

Mk 173 52 ἐν πνεύματι ἀκαθάρτῳ is a Semitism of thought which 

naturally employs this possible construction in Greek. 


(d) ἐν τῷ c. infin. with temporal force. See pp. 25, 450f. A dis- 
tinctive feature of Luke. Hebraic, not Aramaic construction. 


1F. C. B. to J. H. M. (letter, 30/8/10), “οὐκ ἐπιστεύσατε αὐτῷ (Mk 1153 
=Mt 21”) is trd. by the Old Syriac, ‘ believed not in him’: in the Peshitta 
it is ‘ believed not him.’ The Peshitta is literal: the Old Syriac gives the 
Aramaic idiom.” 


SEMITISMS IN THE NEW TESTAMENT. 465 


Ἔμπροσθεν, ἔναντι, ἐναντίον, ἐνώπιον. (See above, § 130.) 

Radermacher (Gr.? 145) speaks of the first and last of these, together 
with πρὸ προσώπου and ἀπὸ προσώπου, as Semitising substitutes in the 
province of mpd. 


Ἔμπροσθεν 84 times in LXX for spd. ᾿Ενώπιον hundreds of times 
in LXX, for either spb or ‘>, ἔναντι and ἐναντίον both very common 
in LXX, generally for ‘Yd, Burney remarks (Aram. Orig. 15) that 
Hebrew distinguishes between "yb in the (physical) sight of, and "ya in 
the (mental) sight of, and finds a corresponding distinction in the NT 
use of ἐνώπιον and ἐναντίον. 

For the three Hebrew expressions, Aramaic uses DP." 

Ἔμπροσθεν is good Greek, surviving in MGr as ἐμπρός, μπροστά. 
The two following uses, however, are undeniably Semitic. 

1. Mt 11°°=Lk 107! οὕτως εὐδοκία ἐγένετο ἔμπροσθέν σου. 
Hebrew ‘25>! i¥2, Aram. DIP NWI=thy good 
pleasure, thy will. 


1 “ ς ᾿ 3 a a 
Mt 1814 οὕτως οὐκ ἔστιν θέλημα ἔμπροσθεν τοῦ πατρὸς ὑμῶν. 
Mt 2318 ὅτι κλείετε τὴν βασιλείαν τ. οὐρ. ἔμπροσθεν τῶν 


ἀνθρώπων, may be due to wrong translation of 
D™P which serves also for ἐναντίον. 


Ἔναντι. For Κοινή use see Wackernagel Hellenistica 1 ff. and 
Vocab. s.v. 
Lk 18, Ac 71° (LXX) 874 NABD (LXX). 


*Evavtiov. Before, in the presence of. For use in this sense in papp. 
see Vocab. s.v. 
Lk 16 2076 2419, Ac 710 ABCD (LXX) 833 (LXX), 


᾿Ενώπιον. See above, p. 15, and Vocab. s.v. Survives in the καθαρεύουσα, 
not in MGr vernacular. [Med. Gr.—R. McK.] 


1. Lk 22 times, Ac 13 times. 

2. Jn 20°, 1 Jn 372, 3 Jn 8, Rev 34 times. 

3, Rom (3), 1 Co (1), 2 Co (3), Gal (1), 1 Ti (6), 2 Ti (2); 4 of 
which from LXX. 

4, Heb (2), Ja (1), 1 Pet (1) 


Katévaytt. Opposite, over against, Mt 213, Mk 11? 124! 13%, 
In the presence of, Mt 274 (BD). 
Rom 42%, 2 Co 217 1219. 
For the former meaning in Kown, see Vocab. s.v. Very 
common in LXX, 


Κατενώπιον. In the presence of, Eph 14, Col 133, Jude 33. 
LXX 7 times (+Dan ©, 1). See Vocab. s.v. 


*anévavne. Over against, Mt 27°1; against, Ac 177. 
In the presence of, Mt 2774 (NAL), Ac 8156 Rom 318 (LXX). 


466 A GRAMMAR OF NEW TESTAMENT GREEK. 


Here we may conveniently take the prepositional phrases with 
πρόσωπον. 


᾿Απὸ προσώπου --Ξ")5}2͵ 


1. Ac 330 541 745, 
2. Rev 618 (? LXX) 1214. 204 (LXX). 
8, 2 Th 19 (LXX). 


Ent mpdcwmov=")5 by 

1. Lk 21°55, - 
᾿Επὶ προσώπου. 

1. Ac 1738 (cf, Jer 3913 [LXX =25% Heb.]). 
Πρὸ προσώπου 355, 


1. Mk 12=Mt 11!°=Lk 727 (LXX). 
Lk 175 (LXX), 9°2 101, Ac 13%, 


In all these passages the influence of the Greek of the LXX is unmis- 
takable. 

Kata πρόσωπον used adverbially, as in Ac 251%, 2 Co 101, Gal 2", is 
certainly not Semitic, but its prepositional use in Lk 231, Ac 313, though 
not uncommon in Greek (cf. Xen. Cyr. vi. iii. 35, τὴν κατὰ πρόσωπον τῆς 
ἀντίας φάλαγγος τάξιν, “ the post immediately in front of the enemy’s 
phalanx’’), is suggested by the OT idiom. 

Radermacher (Gr.? 143) observes that while some of the improper 
prepositions and prepositional substitutes so common in NT are good 
enough Greek (e.g. ἄτερ, ἔναντι, κατέναντι, ἀπέναντι), aS & general rule 
this wealth of substitutes (esp. periphrastic substitutes as ἀπὸ προσώπου, 
πρὸ προσώπου) is due to Semitic influence, from which those papyri 
are not free which exhibit corresponding usages (e.g. ἐνώπιόν τινος). 


Μετά. 


(a) Ποιεῖν (ἔλεος) μετά τινος corresponding to DY TDN ΠΝ (as often 
in LXX, e.g. Gen 2639). 


1. Lk 17? 1087: 158 (μεγαλύνειν ἔλεος μετά). 
Ac 1457 154 (ὅσα ἐποίησεν ὁ θεὸς μετ᾽ αὐτῶν). 
Cf. Tob 12° περὶ ὧν ἐποίησεν μεθ᾽ ὑμῶν. 
1 Mac 1057 καὶ ἀνταποδώσομεν ὑμῖν ἀγαθὰ ἀνθ᾽ ὧν ποιεῖτε μεθ᾽ ἡμῶν. 
Herm. Sim. v. 11 περὶ πάντων ὧν ἐποίησε μετ᾽ ἐμοῦ. 


See Proleg.3 246 and Vocab. 401a, where, on the strength of Kuhring’s 
citation from a Byzantine pap., this “solitary Hebraism left to pera” 
was said to be demolished. But J. H. M. afterwards was inclined to 
withdraw this statement and recognise translation Greek (note by W. F. H., 
1914). For Hebr. and Aram. idiom see Torrey CDA 38) 

(b) Πολεμεῖν peta τινος. 


SEMITISMS IN THE NEW TESTAMENT. 467 


2. Rev 216 127 134 1714 (also ποιῆσαι πόλεμον pera, Rev 117 1217 137 191%), 


See Proleg.? 106, 246, Vocab. 40la for use in papp. and MGr (for 
which see Thumb Hellen. 125, Handb. 103). But in view of Charles’s 
treatment of Rev 127 (ICC i. 322, 356: see above, pp. 448 f.), we should 
probably allow for the influence of the LXX in all these passages. 


Μεταξύ. 


The idiom in Mt 1818 μεταξὺ σοῦ καὶ αὐτοῦ povov=privately is Aramaic 
(W1 32, 226, “‘ unter vier augen,’” Néldeke, Syr. Gr. 189 n.1: “‘ Even 
the Greek text has the Aramaising idiom’’). This is literally rendered 
by syr’®t, which gives the same idiom for κατ᾽ ἰδίαν in 17:9 2017, Mk 9538 
(see W1 32, 226 also McNeile Comm. Matthew, 266). 


Napa. 


Παρά 6. ace. to mark comparison after a positive adj. or a noun, or a 
verb. Semitic acc. to W1 28, 221 (=}1). 


1. Lk 132. 4 1814 (xBL). 


The use of παρά after a comparative is very common in Hellenistic, 
merely extending a classical usage. There are many exx. in NT, esp. in 
Hebrews. But the instances given above conform to the very frequent 
construction in LXX, where Thackeray (Gr. 23) recognises influence of 
Hebr. 1) 7173, but notices that in MGr μεγαλύτερος ἀπό has become 


a normal phrase (see Thumb Handb. 102). 


Πρός. 
Πρός ο. acc.=with has been claimed as due to Aram. mye (=apud, 
παρά, or ad, πρός) by J. R. Harris (OPJ 7 ff.), Burney (Aram. Orig. 28 f.). 


1. Mk 68 (=Mt 1358) 919 (—Lk 941) 1449, 

2. Jn 11,1 Jn 12. 

3. 1 Th 33, 2 Th 25 31°, 1 Co 16% 7, 2 Co 58 119, Gal 118 25 418. 20 Ph 126, 
Philem 13, 

4, Heb 4:38. 


Mr. G. R. Driver (Orig. Lang. 2a) denies Aramaism in any strict sense, 
regarding this as “‘ an extension of many classical usages, particularly 
in such phrases as ἐνθυμεῖσθαι πρὸς αὑτόν. He οἵ. πρός pe παῖσαι “to 
play with me,” quoted by Stephanus (Z'hes. 111. 573). 

Hellenistic usage may be seen in Sy/ll.? 110941 (a.p. 178) διδόντες 
ἡμιφόριον μέχρις ὅτου πρὸς γυναῖκας ὦσιν. Ὁ. 8. Sharp cites Epict. iv, 
9. 13, πρὸς ὃν ἐστί σου πιθανώτερος. With Heb 4138 cf. the formula ὡς 
πρὸς σὲ τοῦ λόγου ἐσόμενου, “ Knowing that you will be held account- 
able,’ P Hib 53% (B.c. 246), P Oxy ix. 118817. [Med. Gr.—R. McK. See 
also Jannaris Gr. § 1658 (c).] 


468 A GRAMMAR OF NEW TESTAMENT GREEK. 


9, CONJUNCTIONS AND PARTICLES. 


"A\Ad.—The Aramaic NDN combines exceptive and adversative 
meanings. Hence it has been claimed that ἀλλά and εἰ μή are some- 
times confused in the NT (W.1 24, #16). We must note, however, that 
in LXX ἀλλά translates ‘Apa in Gen 2159, 

᾿Αλλά for εἰ μή (or ἐὰν μη). 

The clearest cases are: 

v μὴ ἵνα φανερωθῇ" οὐδὲ ἐγένετο 
φανερόν 


Mk 4” οὐ γὰρ ἔστιν τι κρυπτόν, ἐὰ 
ἀπόκρυφον, ἀλλ᾽ ἵνα ἔλθῃ εἰς 
where ἀλλά and ἐὰν μή are parallel ; 


98 οὐκέτι οὐδένα εἶδον ἀλλὰ τὸν Ἰησοῦν μόνον μεθ᾽ ἑαυτῶν (ΑΟΙΥΘ. εἰ 
μὴ NBD 33=Mt 178 εἰ μή sine var.) 


where the alternative texts are equivalent. 

For Mk 1049 (=Mt 2023) see Proleg. 241. In Hinl. 269 n.t, Moulton 
eft. Soph. OT 1331, but adds, “ Of course I have nothing against the recog- 
nition of an Aramaic idiom as the reason for the choice of a similar Greek 
usage to render an Aramaic locution.” 

Ei μή (or ἐὰν μή) for ἀλλά. 

1. Mt 121 ὃ οὐκ ἐξὸν ἢν αὐτῷ φαγεῖν... εἰ μὴ τοῖς ἱερεῦσιν μόνοις. 

Lk 426 καὶ πρὸς οὐδεμίαν αὐτῶν ἐπέμφθη Ἡλείας εἰ μὴ εἰς Sapemta....- 

2. Rev 212? καὶ οὐ μὴ εἰσέλθῃ εἰς αὐτὴν πᾶν κοινὸν καὶ ὁ ποιῶν βδέλυγμα 
καὶ ψεῦδος, εἰ μὴ οἱ γεγραμμένοι ἐν τῷ βιβλίῳ τῆς ζωῆς τοῦ 
᾿Αρνίου. 


8. Gal 3:5 οὐ δικαιοῦται ἄνθρωπος ἐξ ἔργων νόμου ἐὰν μὴ διὰ πίστεως Χ. Ab 


Hort (Comm. James, p. xvi) denies identification in Gal 119: “‘ For 
the very late exchange of εἰ μή and ἀλλά in NT there is no probability 
whatever. In three other books of NT in less good Greek (Mt, Lk, Rev), 
the meaning looks like this, but fallaciously.” 

We may quote for classical usage Xen. Hellen. τι. ii. 10, ἐνόμιζον δὲ 
οὐδεμίαν εἶναι σωτηρίαν εἰ μὴ παθεῖν UPON which G. M. Edwards observes, 
“ες They had no safety except to suffer’ may be equivalent for ‘ they had 
no safety, but must suffer.’ Cf. Mt 26% and Shaks. Kg J ohn Iv. i. 91.” 
(‘‘ Is there no remedy? None but to lose your eyes.”) For ἐὰν pn= 


“but only’? Mr. E. E. Genner eft. Andocides de Myst. 8 89: μηδ᾽ ἐπ 


> ‘ , , dA \ A > A > ‘ ~ > , 
ἀνδρὶ νόμον τιθέναι €av μὴ TOV αὐτὸν ἐπι πασιν Αθηναίοις. 


> 


Et. 
In solemn asseverations εἰ negandi corresponds to Hebrew O8, 


(1 Sam 1445 Sipy ON TON (ἢ Κύριος εἰ πεσεῖται, cf. 2 Sam 1111.) 
1. Mk 813 ἀμὴν λέγω ὑμῖν, εἰ δοθήσεται τῇ γενεᾷ ταύτῃ σημεῖον. 

(N.B.—In parallels Mt 164 1239, Lk 119, καὶ σημεῖον ov δοθησεται.) 
4. Heb 31! 48-5 (LXX). 


SEMITISMS IN THE NEW TESTAMENT. 469 


A Hebrew idiom. ‘‘ Aramaic has nothing like it, except in the 
Targums. Probably a reminiscence of LXX ; Gen 1433 etc.” Lagrange, 
S. Mare, p. \xxxi. 


Kat. 

The use of καί in contrasted statements is a characteristic of 
the Fourth Gospel. Burney (p. 66) treats this as a literal rendering of 
ἡ adversative, which is common to Hebrew and Aramaic. This use of 
καί ““ but” is a slight extension of καί “and yet,” discussed under 
A, 4(f) (8), and can be found in class. Gr., both with and without ov. (See 
K-G ii. 248.) The best examples in the Gospel are Jn 1517". Cf. also 
110. 11 311, 19, 32 420 589. 40. 48, 44 670 74. 19. 30 9290. 52 930 1934 165 2029 9111 
[Med. and MGr.—R. McK.] 

For other uses of καί see under A, 4 above (Parataxis). 


Ὅτι. 

(a) Ὅτι recitativum is of course well established in Greek, 
(Goodwin MT 285f). Guillemard observes that in Gen 2815 4438 [y,1.] it is 
used to translate JS or jas κε verily,” and is often used for ‘5, which 
is sometimes (e.g. Gen 2933, Jos 274, Jer 222) strongly asseverative. He 
claims that in many NT passages (e.g. Mt 7523 107 1476 198 26%. 72. 74 2743) 
that is the force of ὅτι. 

Archd. Allen (Exp T xiii. 330, Comm. Mark 48) attributes frequency 
in Mark to influence of Aramaic 3, For use of ὅτι recitativum in papp. 
see Vocab. s.v. [Med. and MGr.—R. McK.] 


(b) “Orc for relative pronoun. See above B, 1 (e) (8). 


(c) Ὅτι for temporal particle. Burney (Aram. Orig. 78) detects con- 
fusion of  - ὅτι and Ἴ =6re in two passages : 


8 « ~ Dh \ , ef , > 
Jn 98 οἱ θεωροῦντες αὐτὸν τὸ πρότερον ὅτι προσαίτης ἦν. 
124! ταῦτα εἶπεν “Hoaias ὅτι εἶδεν τὴν δόξαν αὐτοῦ. 


In the latter this explanation is unnecessary, whilst in the former 
Mr. G. R. Driver (ut supra) eft. 435, also Ps-Nicod. I. B. i. 3 εἶδον τοὺς 
Ἑβραίους ὅτι ἐστρώννυον ἐν τῇ ὁδῷ Ta ἱμάτια αὐτῶν, Where the ὅτι clause, 
according to the late Greek use, plainly means “‘ strewing their garments 
in the way.” 

Similarly— 


[-} 
Iva. 


(a) “Iva for relative pronoun. See above B, 1 (e) (pp. 434 ff.). 

(b) “Iva for temporal particle. Burney (1.0. 78) accounts for ἵνα as for 
ὅτι above in the phrase ἔρχεται Spa iva c. subj.,—a characteristic of 
Jn (who however uses ὅτε in 421. 23 525 16%, and ἐν 7 in 5?8). 


470 A GRAMMAR OF NEW TESTAMENT GREEK. 


2. Jn 1238 131 162. 821 
[Rev 251, but here iva μετανοήσῃ is clearly final.] 


Mr. G. R. Driver cites several instances from late Greek of καιρὸς 
ἔρχεται (ἐστινὴ ἵνα, whilst MGr ceive καιρὸς va ἔλθης is the regular idiom 
for “‘ it is time for you to come.” To this we may add, ἦρθεν ἡ ὥρα va 
πεθάνῃ, “the hour came to die’? (Thumb Hdb. 187). This usage is 
therefore at most a secondary Semitism, and can quite as easily be 
explained by the writer’s strong partiality for this particle, which had 
already gained great flexibility in the Κοινή. 


“Iva μή. 

Burney (Aram. Orig. 100) finds very cogent proof of translation from 
Aramaic in 
Jn 124°, where iva μή represents the Aram, NPT, though the Hebr. {B 


is rendered μήποτε in LXX, which is retained when the same 
citation (Is 610) is given in Mt 13 and Mk 4%. Moreover, 
‘the Pesh. translates dd. But iva μή is perfectly good Greek 
(see K-G ii. 378 f., Goodwin MT ὃ 315), and its general use 
in the Ptolemaic papyri is clear from Mayser ΟὟ. ii. 240 ff. 
When Burney says that μήποτε never occurs in Jn, though 
found in Mt 8 times, Mk 2, Lk 6, he does not add that iva μή 
is found in every book of the NT except 2 Th, 2 Tim, 1 Pet, 
3 Jn, often with the meaning “lest.” It is equally significant 
that μήποτε does not occur in Rey, though iva μή is found 
there 1] times. Yet Dr. Charles claims that Hebr. not Aram. 
is the background of the Apocalypse. 


Dr. McKinlay observes that in the Modern Version of the NT διὰ va μή 
is used in every passage in. Jn where iva μή occurs. Pallis generally 
has γιὰ va μή, but sometimes μήπως. 


C. MISTRANSLATION OF SEMITIC WORDS OR PHRASES. 


The following instances are amongst those that have been offered : 

1. Mk 25 ἀπεστέγασαν τὴν στέγην καὶ ἐξορύξαντες χαλῶσι. Wellhausen, 
because of the reversed order and the redundancy, retrans- 
lates the first clause SFP, ‘m>pyi which might also mean 
“they brought him to the roof” (17: 37). Schulthess 
(ZNTW xxi, 220) protests that this Aramaic phrase would 
mean “‘ to uncover the roof,” but not “ to bring him to the 
roof,” for which P‘DN would be the right word. 


1C. J. Ball Exp T xxi. 91 so takes it in 8, reading ‘5x for ‘DY 
“rejoiced when he saw my face.” For a different explanation see under 
C, below (p. 475). But see Vocab. s.v. ἵνα (δ). 


SEMITISMS IN THE NEW TESTAMENT. 471 


Μκ 73! ἐξελθὼν ἐκ τ. ὁρίων Τύρου ἦλθεν διὰ Σιδῶνος εἰς τ. θάλασσαν τ. Tar. 
Wellh. (ἰδ. 37) suspects διὰ Σιδῶνος Ξεὺ Τὺ ν 3, which should 
have been rendered πρὸς Βησσαϊδάν (so 6% D). But this 
geographical correction is unnecessary. 

See also Archd. W. C. Allen, Comm. Mark, 50 f. 


Mt 23% καθαρίζετε τὸ ἔξωθεν τοῦ ποτηρίου κ. τῆς mapowidos, ἔσωθεν δὲ 
γέμουσιν ἐξ ἁρπαγῆς x. ἀκρασίας. Luke 1139, rightly, τὸ δὲ 
ἔσωθεν ὑμῶν γέμει. The sense in Mt requires γέμετε for 
γέμουσιν. In Aram., participle in both clauses. (W136, 227.) 

28! ὀψὲ δὲ σαββάτων, τῇ ἐπιφωσκούσῃ εἰς μίαν σαββάτων. Torrey 
(C. Η. Toy Studies, 300) postulates an Aram. ‘73) xnav ‘PIDN2 
Nava IN, “ after the Sabbath, in the night introducing the 
first day of the week.” This is not a case of mistranslation, 
but of a painfully close rendering of an Aramaic phrase. 
We are asked to believe that Mt adds this Aram. clause to 
the sentence in the Greek Mk, which he is following pretty 
closely, and then translates it into Greek ! 


Lk 139 cis πόλιν Ἰούδα. Torrey (1.6. 290 ff.) solves this palpable difficulty 
by supposing Hebr. THN ΠΣ ὃς (better than Aram. 
Sn" TaD), which should have been translated εἰς τ. 
χώραν τῆς Ἰουδαίας. During the first century Α.0. the 
meaning “‘ city’ supplanted the earlier “ province” as the 
force of ΠΣ. 

21 ἀπογράφεσθαι πᾶσαν τὴν οἰκουμένην. Torrey (ἰ.0. 293) conjectures 
PINT", i.e. “all the land” (γῆν), not “ all the world.” 
114! τὰ ἐνόντα δότε ἐλεημοσύνην, καὶ ἰδοὺ πάντα καθαρὰ ὑμῖν ἐστι. 
For the first clause Mt 2986 has καθάρισον πρῶτον τὸ ἐντός, 
which alone gives the right sense. Wellh. (W1 36, 227) 
attributed the variants to a confusion between “5 “ cleanse,” 


and ‘3? “ give alms.” ‘Torrey (/.c. 312) doubts such a use of 
‘Dt in the time of the evangelists, but had independently 
conjectured an Aram. original δ ΡΝ Wy 1355 Nv “ that 
which is within make righteousness.” Here NPTY Ty 


is the regular idiom for “ give alms,” and exactly repre- 
sents δικαιοσύνην ποιεῖν, which has that meaning in Mt 6!-4. 
On the other hand, Burney (Aram. Orig. 9) gives evidence that 
in New Hebr. and Aram. ‘3? means both “to purify” (as well 
as the normal 131) and also “ to give alms.” 

1147: 48 οὐαὶ ὑμῖν, ὅτι οἰκοδομεῖτε τὰ μνημεῖα τ. προφητῶν, οἱ δὲ πατέρες 
ὑμῶν ἀπέκτειναν αὐτούς. "Apa μάρτυρές ἐστε καὶ συνευδοκεῖτε 
τοῖς ἔργοις τῶν πατέρων ὑμῶν" ὅτι αὐτοὶ μὲν ἀπέκτειναν αὐτούς, 
ὑμεῖς δὲ οἰκοδομεῖτε. 


Torrey (l.c. 313) shows that ὑμεῖς δὲ οἰκοδομεῖτε should 


472 


A GRAMMAR OF NEW TESTAMENT GREEK. 


be as Mt 23%! υἱοί ἐστε (αὐτῶν). The Aram. ind 123 PAX 
‘*and ye are children of theirs,” was misread, the noun na 
being easily confused with the participle j!J2 which had 
come in the previous verse, and the word Hind, now taken 


as the direct object, was omitted from the Greek as un- 
necessary. 


Lk 1246 καὶ διχοτομήσει αὐτὸν καὶ TO μέρος αὐτοῦ pera τῶν ἀπίστων θήσει 


(=Mt 2451, where ὑποκριτῶν for ἀπίστων). Torrey (I.c. 314 f.) 
suspects Aram. original NpY DY mn ἜΡΡΕΡΝ “and will 
divide him his portion with the unfaithful,’ assuming that 
ἡ was inserted before "ΠΣ through mistaking the indirect 


object of the first suffix for a direct object. The verb Ὁ" 


would then be added to the sentence to complete the sense, 
and the whole now ran DW” NIpY py Ans ma2dD%, 

On the other hand, the commentators point to the 
barbarous methods of punishing oriental slaves. Plummer 
cites Herodotus for διατέμνω, and Suetonius Calig. xxvii: 
multos honesti ordinis . . . medios serra dissecuit. Moulton, 
Exp T xiv. 430, Vocab. 165, quotes from a sepulchral inscr. 
(iii/iv A.D.) διχοτομέω in a figurative sense. 

In any case Q seems to have been followed by both Mt and 
Lk without question. 


1249 καὶ τί θέλω εἰ ἤδη ἀνήφθη; The required meaning, “‘ How I 


wish that it were already kindled,’ can hardly be got from 
the Greek, but appears at once if the two Hebraisms are 
recognised. (a) 71) not only=ri “ what?” but is used in 


exclamations =‘‘ how,” for exx. see G-K 8148. (b) Θέλω εἰ 
found twice in LXX (unfortunately the underlying Hebr. is 
not available in either passage), “to wish that”: Is 9° καὶ 
θελήσουσιν εἰ ἐγενήθησαν πυρίκαυστοι “ὁ and they shall wish 
that they had been burned with fire” (see Ottley Isaiah i. 
p- 97); Sir 2314 καὶ θελήσεις εἰ μὴ ἐγεννήθης “so shalt thou 
wish that thou hadst not been born.” 

The exclamatory use of τί is found in MGr, τί καλά 
‘* how fine !’’ (Thumb Handb. 181). 

Torrey (l.c. 315) finds that literal translation into Aramaic 
gives the regular idiom for the meaning required NI¥ A193 


ΠΡΡῚ ΛΞ jd PN NIN, 


2482 οὐχὶ ἡ καρδία ἡμῶν καιομένη ἢν; Variants for καιομένη in Ὁ 


and oldest verss. (syr. lat’*t) testify to difficulty felt from 
beginning. Torrey (/.c. 316) suggests 1. “* heavy,” “ slow to 


understand ” for ἼΡ᾽ “ burning” (καρδία = ab = intelligence). 


Ac 


SEMITISMS IN THE NEW TESTAMENT. 473 


247 ὁ δὲ κύριος προσετίθει τοὺς σωζομένους καθ᾽ ἡμέραν ἐπὶ τὸ αὐτό. . 


Torrey (CDA 10 ἢ.) rejects the ordinary meaning of ἐπὶ 
τὸ αὐτό Which is the LXX equivalent for 117‘, pointing out 


that the Aram. word for this is sand “ together,” “‘ into 


one,” and is indeed used in the Pal. Syr. in Jn 1733, and by 
the Syr. verss. in Jn 1153, to translate εἰς ἕν. But in the 
Judean dialects of Aramaic this word means “ greatly,” and 
is used regularly in the Onkelos Targum for the Heb. 180, 


The original Aramaic is thus reconstructed : 817 D119 Sal 
samp of 53 mn vb, The translator is then ‘supposed to 
have misread 1D as a direct object (vide supra, ap. Lk 1246), 
and to have taken the last word 885 --ἐπὶ τὸ αὐτό instead of 


σφόδρα : “‘ And the Lord added greatly day by day to the 
saved.” 


The main objections to this conjecture are: (a) Such a 
blunder is not likely on the part of one who could give the 
right rendering in 67: καὶ ἐπληθύνετο ὁ ἀριθμὸς τῶν μαθητῶν 
ἐν Ἰερουσαλὴμ σφόδρα. (b) A comparison of 1 Co 111% συνερ- 
χομένων ὑμῶν ἐν ἐκκλησίᾳ ἀκούω σχίσματα ἐν ὑμῖν ὑπάρχειν, 
with υ.329 συνερχομένων οὖν ὑμῶν ἐπὶ τὸ αὐτὸ οὐκ ἔστιν κυριακὸν 
δεῖπνον φαγεῖν, shows clearly that ἐπὶ τὸ αὐτό and ἐν ἐκκλησίᾳ 
are Synonymous terms. Indeed, we are inclined to account for 
the juxtaposition of the two phrases in Ac 257 D ἐπὶ τὸ αὐτὸ 
ἐν τῇ ἐκκλησίᾳ as a marginal gloss by a scribe who recognised 
their equivalence. Mr. A. A. Vazakas (JBL xxxvii. 106 ff.) 
shows that the phrase under discussion is an ordinary Greek 
expression very common in the LXX, but in the NT (Ac 118 
21. 47, 1 Co 1118: 29 1423) and the Apostolic Fathers (Barn 41°, 
Ignat. Eph. 131, Magn. 71, Philad. 6? 101, Clem. ad Cor. 34°) it 
has a technical meaning, signifying the union of the Christian 
body. It might generally be translated “‘ in church.’ Professor 
Burkitt (J7\S xx. 321 ff.) repudiates Dr. Torrey’s Aramaic 
explanation and translates ‘“ The Lord was joining such as 
He had foreordained to be saved daily together.” 

J. de Zwaan (Beginnings of Christianity, τ. 11. 55) accepts 
this ‘‘ splendid observation of Torrey.” Dr. Foakes-Jackson 
(HTR x. 358) is forced to the conclusion that “‘ an Aramaic 
original is at the back of this and other strange expressions.” 
Dr. H. J. Cadbury (AJT xxiv. 454), while not granting that 
the intensive explanation of ἐπὶ τὸ αὐτό is really necessary, 
shows cause for thinking that, if it does seem required, “‘ an 
original Hebr. or even an extensive use of the Greek phrase, 
like that familiar to Luke from its abundant use in the LXX, 
would cover the case fully as well.” 


16 Ὁ 5.ὴΐ “ , ~ > , > a - a 6 - \ ἴδ 
9 Kal ἐπι τῇ πιστει του ονοόματος αὐτοῦ TOVTOV ον EWPELTE και OLOATE 


474 A GRAMMAR OF NEW TESTAMENT GREEK, 


ἐστερέωσεν TO ὄνομα αὐτοῦ καὶ ἡ πίστις ἡ δι᾿ αὐτοῦ ἔδωκεν αὐτῷ 
τὴν ὁλοκληρίαν ταύτην ἀπέναντι πάντων ὑμῶν. 

Torrey (.0. 14 ff.) remarks that “‘the ugly repetition of 
TO ὄνομα αὐτοῦ obscures the sense and spoils the sound.” 
A literal rendering into Aramaic is given. 


Tow APN RAIS PYM RAIS PT Tne Ape mT NNT 
sfitdp pp ΝῚ nmyodn ad nam ma 1 smn 
By pointing 72 APN (Ξεἐστερέωσε τὸ ὄνομα αὐτοῦ) as 
my APM (-ε ὑγιῆ ἐποίησεν αὐτόν) the sentence now reads: 
“And by faith in His name He hath made strong this one 
whom ye see and know; yea, the faith which is through Him 
hath given him this soundness before you all.” 
Dr. Burkitt, however, secures excellent sense, as well as 
“ἢ characteristically Lucan rhetoric,” by placing a colon before 
τοῦτον and omitting ἐπί with 8*B. The passage now runs: 
““Ye killed the Prince of Life, whom God raised from the 
dead, whereof we are witnesses, even to the faith in His 
name: this man whom ye see and know His name hath made 


strong, and the faith which is through Him hath given Him 
this perfect soundness before you all.” 


Ac 4% ὁ τοῦ πατρὸς ἡμῶν διὰ πνεύματος ἁγίου στόματος Δαυεὶδ παιδός σου 
εἰπών. 
Torrey (/.c. 16 1.) offers an Aramaic rendering 
PON FAY WT Rvp MT NM DD NBIN ΥἽ NAN 


and shows that the common confusion between x7 and NIN 
has obscured the true meaning, “‘ that which our father 
David, Thy servant, said by the mouth (command) of the 
Holy Spirit.””. The feminine x'7 must be so rendered, but 
the masc. N\n gave rise to the first clause in the Greek text. 
1128 Here, as in Lk 21, Torrey would remove the rhetorical exaggera- 
tion by supposing that an original 8Y7N has been rendered by 
οἰκουμένη instead of by γῆ. In neither case is the explanation 
convincing. The remaining instances given by Professor 
Torrey are less impressive (Ac 810 13! 157). See CDA 18 ἢ, 


and discussions by Burkitt, Vazakas, and de Zwaan already 
referred to. 


2. Burney (Aram. Orig. 103 ff.), in addition to the many examples 
of mistranslated 4, thinks that mistranslation is to be traced 
in the following passages: Jn 115-29 222 683 737. 38 856 925 202. 18, 
But since so eager a supporter of his thesis as Professor 
Torrey confesses that he is “‘ unable to follow Burney in any 
one of these instances” (ΗΤ Καὶ xvi. 329), it will be enough to 
take the two most important. 


SEMITISMS IN THE NEW TESTAMENT. 475 


Jn 787-38 Dr, Rendel Harris (Expos vit. xx. 196) followed some Old 
Latin and Western texts (e.g. D) in redivision of verses, 
which Burney, it seems (1.6. pp. 385 ff.), had already adopted 
for Semitic parallelism. Thus the passage reads: 

Ἔάν τις διψᾷ ἐρχέσθω πρός pe 
καὶ πινέτω O πιστεύων εἰς ἐμέ, 
καθὼς εἶπεν ἡ γραφή, Ποταμοὶ ἐκ τῆς κοιλίας αὐτοῦ 
ῥεύσουσιν ὕδατος ζῶντος. 


{ 


Burney’s solution of this ‘‘ scripture” allusion is gained 
by a blending of Joel 318 (418 Hebr.) and Zech 14°, together 
with an identification of }\Y'>, the word for ‘“* fountain”? in 
Joel (also found in the Targum of Ps 104°, Pr 5'® 825) 
and j‘3!2 (Hebr. D'Y'>) used of “ belly” (Dn 255). 

Mr. G. R. Driver (Orig. Lang. 6 n.) quotes G. B. Gray’s 
criticism, that this explanation ignores the pronoun αὐτοῦ, 
whilst ΠΝ) and A'}'32 would not easily be confused either 
by eye or ear. 

Dr. Rendel Harris’s explanation is drawn from the very 
slight difference between the Syriac words for “ belly” 
Karsd, and for “‘ throne” Kurs’yd. His theory is that in 
an early Aramaic Book of Testimonies a composite quotation 
from Zech 148, “ Living water shall go out of Jerusalem,” 
and Jer 317, ‘‘ They shall call Jerusalem the Lord’s throne” 
occurred, which may also be traced in Rev 22!1’, It is to 
be noted that ὁ θέλων in the last passage corresponds closely 
to ὁ πιστεύων in Jn 738, Under this treatment the words 
καθὼς εἶπεν... ζῶντος are a comment by the evangelist. 

For explanations dispensing with any emendation of the 
Greek text, see W. E. Barnes J7'S xxiii. 421, H. St. J. 
Thackeray, The Septuagint and Jewish Worship, 66 f. 

858 ᾿Αβραὰμ ὁ πατὴρ ὑμῶν ἠγαλλιάσατο ἵνα ἴδῃ τὴν ἡμέραν THY ἐμήν, 
καὶ εἶδεν καὶ ἐχάρη. 

Burney, observing that “‘ rejoiced to see” does not give a 
satisfactory meaning, suggests that “‘longed to see” is what 
we should expect. This he finds in the Syriac s’wah, used in 
the Pesh. for this verb. The ordinary meaning of s°wah is 
“long for,’ but in Mt 1218 “ delights in,” “‘ rejoices,’ where 
in the quotation from Is 421 it represents the Hebr. 137, 


He acknowledges that the verb is not known to occur in 
W. Aramaic. 

Torrey (HTR xvi. 340) improves on this by showing that 
in the Targums Ya (¥33) is the regular equivalent of the 
Hebr. bys and the Gr. ἀγαλλιᾶσθαι, whilst the most common 
Aramaic verb for “seek,” “ pray,” is ΟΞ, It is easy to see 
how the final δὲ might be dropped by haplography, ὉΠ Σὲ Nya 

VOL. II. PART III.—32 


476 


A GRAMMAR OF NEW TESTAMENT GREEK. 


““Abraham prayed that he might see,’ thus becoming 
ONAN ya “ Abraham exulted to see.” Lagrange seems to 


doubt whether the form Y3 is found. In any case Burney’s 


suspicion of an Aramaism in the iva clause is needless. 

Dr. McKinlay quotes ἀγιαλλοῦται ἡ ψυχὴ va μάθῃ διὰ τὴν 
κόρην from a mediaeval romance, and cft. MGr χαίρομαι νά 
σε θωρῶ, “I'm glad to see you.” 

It should be added that Torrey proposes (l.c. 338 ff.) 
emendations of the text by working back to an Aramaic 
original at 75-87 1135-88 142. 31 2017, 


Archdeacon Charles has shown with abundance of illus- 
tration throughout his Commentary on the Revelation of 
John that the successful resolution of difficulties in the text 
lies often in retranslation into Hebrew. The following 
examples are the most impressive : 


Rev 10! καὶ οἱ πόδες αὐτοῦ ὡς στῦλοι πυρός. 


Hebr. 227, which normally =“ foot,” also means “leg” 


(so rendered by LXX in 1 Sam 178, Dt 2857), whilst πόδες 
also translates D'Y1D =“ legs,” “ thighs,” Ex 2917, and other 
passages. 

This is obviously the meaning here. (See Charles Studies 
97 ff., ICC i. 259.) [So often in Med. and MGr.—R. McK.] 

133 καὶ ἐθαυμάσθη ὅλη ἡ γῆ ὀπίσω τοῦ θηρίου. 

That is TN “INND PINT-PD FNM), where “INND is 
corrupt for NIN --βλέπουσα. Cf. 178 καὶ θαυμασθήσονται οἱ 
κατοικοῦντες ἐπὶ τῆς γῆς, ὧν ov γέγραπται τὸ ὄνομα ἐπὶ TO 
βιβλίον τῆς ζωῆς ἀπὸ καταβολῆς κόσμου βλεπόντων τὸ θηρίον. 
Also 175 ἐθαύμασα ἰδὼν αὐτήν. (See ICC i. 351.) 

1311 καὶ ἐλάλει ὡς δράκων Ξε 373 723M) probably for IAD IANA) 
=kal ἀπώλλυε Or kai ἦν ἀπολλύων. This very confusion occurs 
in 2 Chr 2910 ΠΞΟΟΙΘΠ ΟΞ ΤΣ IIA) OPM (LXX ἀπώλεσε, 
Vulg. interfecit, scribal error for JANA) 2 ΚΙ 111, Oxf. Hebr. 
Lex. 1816). Cf. Rev 91! ὄνομα αὐτῷ βραιστί ᾿Αβαδδών, καὶ ἐν 
τῇ Ἑ) λληνικῇ ὄνομα ἔχει ᾿Απολλύων. (See Charles Studies 
100 f., ICC i. p. cli.) 

15° (nvotyn) 6 ναὸς τῆς σκηνῆς τοῦ μαρτυρίου ἐν τῷ οὐρανῷ. 

This very difficult phrase = ΒΞ ἽΝ brik bon which 
may be a corruption of powaw ΤῊΝ “Ss = ὁ ναὸς τοῦ θεοῦ 
ὁ ἐν τῷ οὐρανῷ. (Cf. 1119 τοῖς this very sentence occurs, 
ἠνοίγη ὁ ναὸς τ. θεοῦ ὁ ἐν τ. ovp.) 

165 ἐνδεδυμένοι λίθον καθαρὸν λαμπρὸν. 

Λίθον AC (defended as original text against. λίνον by WH 
App.* 139) is extremely difficult. Charles (JCC ii. 38) shows 


SEMITISMS IN THE NEW TESTAMENT. ATT 


that ἐνδεδυμένοι λίθον -ὖν Ὁ 30, But Ww) = βύσσινος in 
Gen 4142, Ex 2835, whilst it seems to mean λίθος in Esth 1, 
and the phrase which is there rendered στῦλοι λίθινοι 
(ww “T32P) is given as στῦλοι μαρμάρινοι in Cant 51. For the 
same mistranslation as here in Ep Jer see Dr C. J. Ball’s 
note OA. i. p. 610. 

(272 ἰδοὺ βάλλω αὐτὴν εἰς κλίνην = ravine rns 5.52 237, where βάλλω 


represents the causative of πίπτω and aavinb ‘bp =" fall ill.” 


See Charles Studies 99, [CC i. 71. Here retroversion does 
not remove a mistranslation, but explains an obscure idiom.) 
[But the pass. of Ba\\o=“ to be laid up” occurs in Aesop 
Fab. cclvii (cited Field Notes 70 n.3); πίπτω in Med. Gr.= 
“to fall ill,’ and in MGr répra=“to go to bed,” “ fall 
asleep.” —R. McK.] 


D. GENERAL SUMMARY. 
1. CLASSIFICATION OF SEMITISMS. 


In the Introduction to this volume (pp. 14 ff.) Semitism has been 
defined as “‘ a deviation from genuine Greek idiom to a too literal render- 
ing of the language of a Semitic original.” The term secondary Semitism 
marks a possible but unidiomatic Greek construction, which strains 
ordinary Greek usage to conform to a normal Semitic construction. 
Dr. J. de Zwaan! has analysed the possibilities very carefully by dis- 
tinguishing between (a) a ‘‘ Greek,” i.e. a writer with a perfect knowledge 
of Greek, and (6) a‘‘ Semite,”’ 1.6. one who writes Greek with an imperfect 
knowledge of the language. He then shows that either (a) or (0) may 
attempt four things: translation from a Semitic dialect into (1) idio- 
matic, or (2) Semiticising Greek ; original composition in (3) idiomatic, 
or (4) Semiticising Greek. Primary Semitisms are said to be those 
which a ‘“‘ Semite ’ commits in attempting (1) and (3), though his im- 
perfect knowledge of natural Greek may betray him in (2) and (4). 
Secondary Semitisms are said to be those deviations which a “‘ Greek” 
may let pass in cases (1), (2) and (4) “‘ through such factors as the exi- 
gencies of his readers or of the documents he is translating.’ Another 
factor is the degree of familiarity which this ‘‘ Greek ” translator has with 
the Semitic idiom of the source which he is using. This subjective 
aspect is important, but in many of the points in dispute we cannot say 
whether the author, or the translator of the source which he is following, 
is properly to be designated (a) or (ὁ). We start from the objective 
data and consider them in their double relationship to Semitic usage, 
and to the tendencies that appear in the later stages of the development 
of the Greek language. When all allowance has been made for the 
coincidence of many Semitic constructions with those found in Hellen- 
istic, the presence of a large number of apparent Semitisms in any writing 


1 Beginnings of Christianity, τ. ii. 53 ff. 


478 A GRAMMAR OF NEW TESTAMENT GREEK. 


will justify the application of de Zwaan’s term ‘‘ Semiticising Greek.’ 
But what criteria will enable us to determine whether this ‘‘ Semiticising 
Greek ” is due to literal translation, or has found its way into original 
composition in Greek through inadvertence, or habit, or design? Pro- 
fessor Torrey 1 mentions three methods by which the fact of translation 
can be demonstrated in a document which resembles an original com- 
position. (1) The precarious evidence of the feeling that certain phrases 
and constructions “‘ sound Semitic rather than Greek.” (2) Mistransla- 
tion, ?.e. the removal of a grave difficulty in the Greek text by the restora- 
tion of the Semitic original, which was thus evidently misunderstood or 
misread by a translator. (3) The cumulative argument based on “ the 
continual presence, in texts of considerable extent, of a Semitic idiom 
underlying the Greek.” But obviously the first and third of these tests 
will not enable us to distinguish between (2) and (4) in de Zwaan’s classi- 
fication, whether the writer be a ‘“‘ Semite” or a “‘ Greek.” Even a 
translator may by force of habit introduce Semitisms into his Greek when 
the Hebrew original is quite different, as many examples from the LXX 
would show. But when once the LXX had become a standard of sacred 
speech for Hellenistic Jews and proselytes, its idioms would easily find 
their way into free composition. Dr. H. J. Cadbury? very aptly uses 
the analogy of extempore prayer in public worship, and the use of Biblical 
language by Bunyan and Lincoln. Torrey writes of the second test, 
that it is ““ immensely valuable in the rare cases where it is convincing : 
there is no other internal proof of translation which is so immediately 
cogent.” Certainly it is a method to be applied with caution, for even 
Dr. Torrey confesses that “‘ it happens in nine cases out of ten that renewed 
study of the ‘ mistranslations ’ which we have discovered shows us that 
there was no translation at all, or else that it was quite correct.”* A 
striking example of this difficulty is furnished by the Acts of Thomas. 
This work survives in both a Greek and a Syriac text. Here there can 
be no question that one of these is a translation of the other. But even 
with the two texts available for comparison, Professor Burkitt finds most 
of the more obvious lines of argument ‘‘ double-edged,’’ and points to 
the ““ ambiguous nature of much that might have been expected to pro- 
duce results.” ‘‘ The only way by which we can prove the Greek to be 
taken from the Syriac is to find instances where the Greek translator has 
actually mistranslated a Syriac idiom, or has followed a reading which 
rests upon a paleographical corruption in the Syriac.” 4 

It is for this reason that so much importance attaches to such instances 
of possible mistranslation as are collected in C above. Even here, 
however, we must observe that the Semitists themselves are not in 
agreement. Dr. Burney, in face of Wellhausen’s well-marshalled evidence, 
declares with regard to Mark, “ What is needed to substantiate the theory 


1C. H. Toy Studies, 283 ff. 

2 AJT xxiv. 453. An instructive contrast can be drawn between the 
simple Biblical English of John Wesley’s published sermons (see p. 9) and 
the crisp conversational English of his Letters and Journals. 

* CO. H. Toy Studies, 284. 4 JTS i. 282. 


SEMITISMS IN THE NEW TESTAMENT. 479 


of an Aramaic original is some cogent evidence of mistranslation ; and this 
has not yet been advanced.” ἢ In the same way Dr. Torrey, who promises 
to produce such evidence of mistranslation in Mark as Wellhausen failed 
to give, thinks that the case for John is weaker. “ Burney’s argument, 
for all its learning and acumen, weakens at the crucial point. Among 
those who are inclined to demand in John what Burney demands in Mark, 
I think the verdict is likely to be ‘ Not proven.’ ” ? 

Even when there is the strongest reason to suspect a translator’s 
error, we are often left in doubt whether this is due to a corruption in 
the original document, to a mistranslation of the originai text, or to a 
linguistic confusion in the writer's mind with no documentary cause at 
all. Thus in the notorious example referred to by Mr. G. R. Driver, did 
the translator of Mommsen faithfully reproduce a printer’s blunder 
Feuerwerk for Feuerwehr, or did he, by some inexplicable confusion, 
translate Feuerwehr by fireworks ? We know that the book is a transla- 
tion, and that the context requires fire-brigade. Only a reference to the 
original German edition can settle the point.* But such errors occur 
when there is no documentary explanation. I once heard so perfect a 
bilinguist as Mr. Hilaire Belloc in a lecture on the French Revolution 
speak of the “‘ sermon in the tennis court.” Though the speaker in- 
stantly corrected himself, the audience could recognise at once the con- 
fusion between serment (oath) and the other French word indistinguish- 
able in sound. There is no more brilliant conjecture in Wellhausen’s 
work on the Gospels than his solution of the difficult ra ἐνόντα δότε 
ἐλεημοσύνην (Lk 1141). The sense requires καθάρισον, which is actually 
found in the Matthaean parallel (Mt 23°), and, as we have seen above 
(p. 471), Welihausen makes this a moral certainty by restoring the 
Aramaic. But what inference are we to draw? There are three alterna- 
tives. (a) Mt and Lk may both have translated from a common Aramaic 
original, one correctly, the other incorrectly ; (Ὁ) Mt and Lk may have 
had the same Greek translation, but whilst Mt knew enough Aramaic to 
correct the mistranslation, Lk faithfully retained it ; (c) Mt and Lk may 
have used different editions of a Greek translation of Q. Similarly, if 
we are convinced that some of the idioms in the Fourth Gospel pre- 
suppose Aramaic, we have still the further point to settle, whether there 
ever was an Aramaic Fourth Gospel, elsewhere than in the mind of the 
author who wrote directly in Greek. In other words, granted that 
‘“« John’ was a ‘‘ Semite,” to which of de Zwaan’s four classes does the 
Greek Gospel belong ? 


2. Tue Semitic Stamp IN TRANSLATION GREEK. 


We have seen that Professor Torrey emphasises the importance as 
also the precariousness of his first test. It is the starting-point in any 
investigation. Nor is it entirely subjective, for it can be applied, espe- 
cially where Hebraisms are concerned, by watching the tendencies of 


1 Aram. Orig. 19. 2 ATR xvi. 332. 
3 [ have failed to discover the passage. 


480 A GRAMMAR OF NEW TESTAMENT GREEK. 


translators when we know Hebrew originals to underlie the Greek. The 
comparative study of the LXX with the Hebrew text is invaluable for 
this purpose, and helps to furnish the list of suspected Semitisms pro- 
vided by the Oxford Apocrypha in the critical introductions to several 
of the books. Still more valuable is the evidence which Canon Box 
gives in his edition of the Ezra-Apocalypse to show that the Latin text 
preserves a number of Hebrew constructions mediated through a lost 
Greek text. But if the presence of many such constructions in any one 
book raises a presumption of translation (or else suggests a deliberate 
adoption of Semiticising Greek), the absence of such constructions will 
tell against translation. The Semitic mind of the writer may, however, 
sometimes betray itself by the repetition of a solitary Semitism,! such as 
the adverbial use of προστίθημι in Josephus (see p. 445). Generally 
speaking, the presence of numerous Hebraisms will suggest the influence 
of the LX X, whereas numerous Aramaisms or idioms common to Hebrew 
and Aramaic will point to a background of Aramaic. With the ex- 
ception of parts of the Apocalypse and the first two chapters of Luke, 
it seems hardly likely that Hebrew sources were translated by any of 
the NT writers. 


3, CoNDITIONS UNDER WHICH THE GOSPEL TRADITION BECAME FIXED. 


The four Gospels record events which happened on the soil of Palestine, 
and preserve the sayings of One whose mother tongue was Aramaic. 
The primitive Christian tradition was inevitably formulated and trans- 
mitted with an Aramaic colouring. The second stage of the Christian 
movement had Antioch as its headquarters, a bilingual city where this 
tradition passed over into its Greek form. It was only after evangelists 
and teachers had carried the Christian message into the world of Hellenism 
that the paradosis was stereotyped in documents. It is hardly open to 
question that our first and third evangelists drew material from docu- 
mentary sources, and it is almost certain that these were written in Greek. 
A factor often overlooked in discussions of the Semitic tinge of the Gospels 
is the linguistic miliew in which the authoritative tradition grew up. 
Its pre-documentary history lies in a region and a community where 
there would be a tendency to fit the idioms of the Kowy as closely as 
possible to the Aramaic tradition.2 This was a community in which 
the OT was not only studied in its authoritative Greek translation, but 
widely used in the form of V'estimonia, compiled first in Aramaic, then 
translated into Greek. The leaders in those formative years when the 
Church was acquiring a Christian vocabulary and phraseology were men 
whose habits of thought were Jewish. It is thus not surprising that we 


1 Τ recall a German friend who spent all his boyhood in London, and 
speaks perfect English but for the German idiom, “ I should like to go with” 
(Ich méchte gern mitgehen, omitting the unaccented pronoun “ you’’). 

* See Schulthess Das Problem, pp. 43, 56. 

3 J. Rendel Harris Testimonies, i. 125. 


SEMITISMS IN THE NEW TESTAMENT. 481 


find ‘‘ Semitisms of Vocabulary ” 1 in the Pauline letters and in Hebrews, 
where grammatical Semitisms are very rare and purely “ secondary.” 


4. THe SEVERAL Books. 
1. Synoptic Gospels and Acts. 


Mark is the most Aramaic of the Gospels. The very few Hebraisms 
may be attributed to the use of the LXX in the Gentile mission, but they 
are echoes few and faint. Burney’s misgivings have already increased 
the doubtfulness of Archdeacon Allen’s claim that the earliest of our 
Gospels was originally written in Aramaic. When Dr. Torrey’s promised 
list of mistranslations appears, a good case may be made out for the 
partial use of Aramaic memoranda. But so far we remain of the opinion 
so well expressed by Pére Lagrange, ‘‘ His Greek is always Greek, yet 
translation Greek; not that he translates an Aramaic writing, but be- 
cause he reproduces an Aramaic κατήχησις.᾽ 

When we turn to Matthew it is perplexing to find these two eminent 
Aramaists in reverse disagreement, for while Dr. Allen bases his belief, 
with most scholars, on the demonstrable use of a Greek Mk by the author 
of the first Gospel, the French commentator argues that our Mt is a trans- 
lation from an Aramaic original. This is perhaps largely due to the 
great stress which the Roman Catholic scholar puts upon early Church 
tradition. But his masterly array of the linguistic data is not convincing, 
especially when one observes how many of the Aramaisms are found in 
the teaching of Jesus. It is just in this part of the Gospel that we should 
look for the Aramaic idiom to emerge, however free the author’s Greek 
elsewhere. There is one point, however, on which the present writer 
thinks that fresh evidence may modify the judgment passed by Dr. 
Moulton.2 Dr. Burney’s remarkable study of the parallelism of the say- 
ings of Jesus gives weighty support to Harnack’s preference for the 
Matthaean as the more faithful record of the discourses.* Though this 
is a matter of poetic structure rather than of syntax, it may well be urged 
that the author of the first Gospel was familiar with Aramaic, and recog- 
nised the Semitic form behind the Greek rendering that lay before him. 

On two points only is it needful to add anything to what has been 
written above (pp. 18 ff.), regarding the Lucan books ;* for the Infancy 
narratives, and more especially the hymns in Lk 1-2, constitute a special 
problem, and the theory that we have a translation of an Aramaic docu- 


1 See Proleg. 11 f., also p. 26 above. 

2 See pp. 10 and 20 above. But in view of Burney’s argument it is the 
more important to re-read Moulton’s two Expositor articles referred to on 
p. 10 n.*. 

3.366 The Poetry of our Lord, 7. 

4 As a footnote to Moulton’s judgment from the Hellenistic side that 
Luke probably did not speak Aramaic, we add the opinion of a distinguished 
Aramaist, “‘ Il n’était pas juif de naissance ni d’éducation, et s'il était Syrien 
d’origine, rien ne prouve que l’araméen ait été sa langue maternelle” (La- 
grange 5. Luc xcvi). 


482 A GRAMMAR OF NEW TESTAMENT GREEK. 


ment throughout Ac 1-15 has entered on a new phase since Dr. Moulton’s 
Introduction was left unfinished in 1915. 

It may be well to refer to Harnack’s thorough investigation of the 
linguistic phenomena in Lk 1-2,} and to quote his emphatic judgment : 
‘“The vocabulary and style characteristic of Lk 1-2 are so absolutely 
Lucan that, in spite of all conjectures that have been made, the hypo- 
thesis of a Greek source is impossible, for there is almost nothing left 
for it. Two things only are possible: either Luke has here translated 
an Aramaic source, or he was dependent for his subject-matter upon no 
written source at all, but has followed oral tradition, with which he has 
dealt quite freely, so far as form is concerned. At all events the two 
great psalms of Lk 1-2 were not handed down to the author (either in 
Greek or Aramaic), but were composed by himself.” “‘ It is possible that 
for the narrative an Aramaic source has been used, but this hypothesis 
is not probable. On closer view the Magnificat and the Benedictus 
present the form of a single complicated, correctly constructed Greek 
period that does all honour to the author of the prologue. This period 
is simply forced into its Hebrew dress. The hands are Esau’s hands, 
but the voice is that of Jacob. But if this is so, then it is plain that Luke 
in composing these canticles has purposely kept to the language of the 
Psalms and prophets (LXX). The Hebraisms, whether adopted or 
inserted from the Old Testament, are intentional; the whole style is 
artificial, and is intended to produce an impression of antiquity.” Now 
Professor Torrey? will not hear of such deliberate imitation of the 
language of the LXX, “‘ for the motive for such a grotesque performance 
on his partis by no means apparent.’ But heis as convinced as is Harnack 
that ‘* the Gospel of the Infancy ” is by every consideration of vocabulary 
and style the language of Luke himself. In his judgment the only 
satisfactory theory is “‘ that the author of the Third Gospel himself 
translated the Narrative of the Infancy from Hebrew into Greek.” 
The strongest argument for translation from Hebrew, either by Luke 
or by the translator of his source, is the use of a phrase which does not 
occur in the LX X and is yet a “‘ translation of the painfully literal kind.” 
The example given by Dr. Torrey is Lk 151, ἐποίησε κράτος ἐν βραχίονι 
αὐτοῦ, obviously a rendering of 113 ben nvy “or (less probably) its 
Aramaic equivalent.’ But if we examine this verse with care its diction 
can easily be paralleled from the LXX. Thus, Lk 181: 


ἐποίησεν κράτος ev βραχίονι αὐτοῦ, 
διεσκόρπισεν ὑπερηφάνους διανοίᾳ καρδίας αὐτῶν. 
ΘΕ ἘΞ 58:0": 
eg) , € , e ° 
σὺ ἐταπείνωσας ws τραυματίαν ὑπερήφανον, 
Nees. “ ,, a ΄ , , \ > , 
καὶ ἐν τῷ βραχίονι τῆς δυνάμεώς σου διεσκόρπισας τοὺς ἐχθρούς σου, 
and: 111: 
δεξιὰ Κυρίου ἐποίησεν δύναμιν. 


1 Luke the Physician, 96-105, 199-218. 
2 See C. H. Toy Studies, 286, 295. 


SEMITISMS IN THE NEW TESTAMENT. 483 


The Hebraic phraseology is beyond question, but there is nothing that 
lies beyond the range of composition by one who was steeped in the 
diction of the Greek version of the Psalter.’ 

The theory that the first fifteen chapters of Acts are Luke’s literal 
translation of an Aramaic document has been urged with a mass of 
detailed illustration by Professor Torrey, who claims that his argument 
is cumulative. We have already considered some of his strongest in- 
stances of possible mistranslation, and can only mention here one or two 
reasons for the failure of this great Semitist to win general consent to 
his conclusions. Several of the alleged Aramaisms can be paralleled 
from Paul’s letters and other writings in free Greek. Others are found in 
the second part of Acts, where they are attributed to the influence of 
the LXX on the writer’s style. Others can be paralleled in Luke’s 
Gospel, but not from Mt and Mk, which are declared by Dr. Torrey to be 
translated directly from the Aramaic. Finally, instances are given of 
mistranslation although in the same part of the book the correct phrase 
is found, thus proving that ignorance of the true meaning of the idiom 
must not be imputed to the author.* 


2. The Johannine Writings. 


The case for an Aramaic original of our Fourth Gospel has been 
greatly strengthened since the editor wrote the sentences on this subject 
in the Introduction to this volume (p. 32). It was but natural to rely 
upon the great authority of Wellhausen, as Dr. Torrey* has generously 
allowed. But Dr. Burney’s masterly work has proved convincingly 
the Semitic cast of mind of the author. His attempt to prove that 
the Gospel was written in Aramaic by one man, and translated into 
Greek by another, has not carried the same conviction. The decisive 
factor in the establishment of such a theory is a few instances of almost 
certain mistranslation. Now, as we have seen above (p. 474), Dr. Burney 
has furnished us with several plausible examples, but it is significant 
that such highly competent Semitists as Pére Lagrange and Mr. G. R. 
Driver recognise that even the most difficult passage can be understood 
without recourse to this kind of treatment. Moreover, the most impres- 
sive evidence which is offered for the confusion of the particles ἵνα and ὅτι 
with the relative loses much of its force when this same tendency is found 
to be increasingly prevalent in the later stages of the Greek language. 
The weakness of Dr. Burney’s case is that he has not allowed for the 
coincidence of many of the Aramaic constructions found in the Greek 
of the Fourth Gospel with usages that were equally common in the 
colloquial Greek of that period. It is only to be expected that one whose 


1 A striking feature of the Third Gospel is, to use a phrase of Lagrange’s, 
the “nests of Semitisms’” which we meet with here and there in Lk. 

2 All these statements are exemplified in Dr. H. J. Cadbury’s searching 
examination of Professor Torrey’s contention: see Luke: Translator or 
Author ἢ (AJT xxiv. 436-455). 

3 HTR xvi. 324. 


484 A GRAMMAR OF NEW TESTAMENT GREEK. 


native tongue was Aramaic would tend to fall into those forms of speech 
when writing Greek which most closely resembled his own idiom. 
Again, when we notice how many of the Aramaisms are found in passages 
that profess to record the actual words of Jesus, we may well suppose 
that John ‘‘ was mentally translating, as he wrote, logia handed down 
by tradition and current in Christian circles in Aramaic, from that 
language into the Greek in which he was actually composing his Gospel.” * 
Finaily, it is only right to point out that Dr. Torrey, who is entirely 
with Burney in his main contention, disagrees with all his attempts 
at the recovery of a mistranslated original. The same might almost 
be said of Dalman’s verdict.2. A written Aramaic original is therefore 
by no means established. 

The Hebraic style of many passages in Revelation was made clear 
by Archdeacon Charles in his Studies in the Apocalypse (see p. 33 above). 
Since the closing page of the Introduction to the present volume was 
written nine years ago, Dr. Charles’s incomparable edition of the Apoca- 
lypse has appeared, with a wealth of material for those who study the 
grammar of this book. 

The solution of the tangled problem of the language of the Apocalypse 
is said to be this: (a) The author writes in Greek, but thinks in Hebrew ; 
(b) he has taken over some Greek sources already translated from the 
Hebrew; (c) he has himself translated and adapted some Hebrew 
sources. The instances of mistranslation corrected by retroversion 
which have been given above go some way to proving the third state- 
ment, though they might still better come under the second heading. 
One ventures to wonder whether the first assertion has been made good. 
The writer’s familiarity with Hebrew seems to lie beyond question, 
but why should not Aramaic be his mother-tongue, the language in 
which his thoughts would first frame themselves? Many of the 
peculiarities of idiom, such as the use of the resumptive pronoun after 
a relative, the co-ordination of a participle with a finite verb, the casus 
pendens, might betray an Aramaic cast of sentence. It is also note- 
worthy that several of the instances of the very free use of iva in the 
Fourth Gospel can be paralleled in Revelation. All of these come 
within the range of late Greek usage, and show that the writer was more 
familiar with the vernacular than with literary models, and naturally 
adopted such locutions as he found most in accord with his Semitic 
habit of speech. But we are convinced that more importance should 
be allowed to the influence of the LXX. One instance must suffice. 
We have already referred to Dr. Charles’s acute perception of the Hebraic 
idiom behind the crux in 127. But why should δ c. inf. in Hebr. be 


rendered by τοῦ c. inf. by one who has to give a desperately literal transla- 


1G. R. Driver The Original Language of the Fourth Gospel, 1 n. (This 
reprint from The Jewish Guardian is the most complete and competent 
criticism of Dr. Burney’s thesis that has yet appeared.) 

2 ThLZ, xviii. 8. ‘‘ Die angenommenen Uebersetzungsfehler, von denen 
oben nur eine Auswahl mitgeteilt wurde, sind nicht zwingender Natur.” 


SEMITISMS IN THE NEW TESTAMENT. 485 


tion? He might have used εἰς τό c. inf. Dr. Charles himself shows that 
at Hos 9:3 the LXX translates the same idiom in precisely the same way. 
Is it not likely that one who was trying to write in Greek, a language 
with which he was not perfectly familiar, would prepare himself for 
the sacred task of declaring his heavenly message by studying the 
revelation of bygone seers, not only in the sacred tongue of the Hebrews, 
but also in the version which was hallowed as the Bible of the Greek- 
speaking Dispersion and of the Gentile Christian Church? We there- 
fore think that the material supplied in Dr. Charles’s great commentary 
would justify us in finding a solution of the linguistic problem in a 
combination of factors: (a) a mind that thought in Aramaic and found 
in the vernacular Greek of his world many idioms sufficiently close to 
his mother-tongue for his purpose ; (δ) sources in translated Greek and 
in Hebrew, which he worked into his book in Hebraic Greek; (c) a 
knowledge of the LXX and of various apocalypses already current in 
a Greek form, which supplied him with a vocabulary and often sug- 
gested an idiom.? 


Of the remaining books of the New Testament there is little to add 
to what has already been said in the Introduction. Those who think 
that Paul’s amanuensis or colleague, who was allowed a freer hand in 
the composition of Ephesians than of any other letter, was deeply under 
the Hebraic influence of the LX X, will find a few additions to the data 
offered on p. 22. Secondary Hebraisms may be discovered in Eph 1% 
εὐλογήσας ἐν πάσῃ evdoyia. Apart from the use of ἐν, we are reminded 
of the group of pleonasms in this Epistle to which Dr. Rendel Harris 
has called attention (p. 419). An unidiomatic use of the genitive of 
definition may perhaps be termed the Hebraic genitive in 144 τῷ πνεύματι 
τῆς ἐπαγγελίας (“the promised Spirit”) and 4% ras ἐπιθυμίας τῆς 
ἀπάτης (“deceitful lusts”). The breathless sentence which spans the 
second half of chapter 1 is built after no Hebrew model, yet a keen eye 
may possibly detect Dr. Charles’s construction (see pp. 34 and 429) 
in 17% ἐγείρας αὐτὸν ... καὶ καθίσας... καὶ πάντα ὑπέταξεν. The 
next line shows us ἔδωκεν (clearly jN), cf. 411). Yet another secondary 
Hebraism may lurk in the phrase (61%) ἐν ἀνοίξει τοῦ στόματός μου 
(2 c. infin. “when I open... .’’). We have already seen (p. 453) that 


exegesis is against treating the participle as otiose in 217 ἐλθὼν εὐηγγελίσατο. 
Nor does there seem good reason for agreeing with Guillemard that 
μαρτύρομαι ἐν Κυρίῳ (417) is the familiar 2 Yavin, Rather is it analogous 
to Paul’s παρακαλοῦμεν ἐν κυρίῳ ᾿Ιησοῦ (1 Th 41) and to the words in 
Rom 167%, with which another Tertius would now bring his pious duty to 
a close, ἀσπάζομαι ὑμᾶς ἐγὼ Τέρτιος 6 γράψας τὴν ἐπιστολὴν ἐν Κυρίῳ. 


1 For further exx. of τοῦ ο. inf. as common LXX equivalent for 5 c. inf. 
in Hebr., see p. 449 above. 

2 For a rather different explanation of the language of the Apocalypse, 
see Lohmeyer’s recent commentary, HNT tv. iv. 193 ff. 


a 


— Ά 
Aigo dt Αἰ σατο αὐνηθῆον. 
᾿ WS Unto a: wie: ye ple 
Wat e- tka plata. νυ; tA. mies Yur τς 
᾿νθχενί! κἂν en i tie ea 
Siweriyo=th Sore erode ai Be hah τῳ 
Pe a 7 é pins es ὄν i ΩΣ εἶν δ if Ss δὲ "- ὌΨΑ ᾿ 
~~ ᾿ Ξῳω oe oat ϑνν Li betigh ow, tee ἢ reek seh 


1 sion yaaa Ἀν sped aglitae ays ee wee’ wk 


Sun ΠΟ Conary ἀρ σύψακ ty phar Ὡς, en teRd ~ τῇ Sti 
+: [- τὸ at oi wh wed tt als ΔΝ; νὸν; “aah πῶ Ὧν 7) ii 


Εν 


- « a 
‘ - i o> fame γῇ bd 
3 oo 
4 . - . we 
2 be " ᾿ 
."᾿ 4 
rey ’ 
a i] ᾿ i ‘ AZ 
was $3 ' 8 Pere τὺ > 
. fee 2 4 * δ΄. ἀλ ὃς ὡ- Gh os ‘ 
ba « 
wi πῶ ν 
πὶ ἂν Μ Ys 3) ὴ Phd , 4 Ἂς 
᾿ Η i ἢ ὁ 2 γε 
eee Ps 
afi Ἂ ἃ p 5 ve ee 
4 « - « oh D6, 4 : 
. ῃ ty 
4 i 
ΡΞ 2 ΜΙ 
ako ἘΝ, Ά ’ é : ad > Ἰ 
+ - ᾿ ~ ω 
δι = “ 
7 . ‘ ῶ ν - 
. oat - 7 
4 ‘ 
sy J.¢ ‘ = 
h 4 ᾿ 
f ᾿ 
“we ὡς 
af 
4 « ‘ Ω 
7 = ΝΜ ~ ee ‘ 
“αν ὦ 
ν - . 
᾿ ὼ a) . 
ἔν: i τὸ 
‘ e . 
- - > + 
A . ’ ‘ 
<a & } 
aru » δὰ ae 
> He + ᾿ Τ - ΨΥ 
2 ᾿ 5 eo "Δ: 
{πὰ κὰν υἱὲ me : νὴ 
ἱ Σ ὦ ἀπ Ὁ} * ’ - ᾿ 
7 
< > οἴ οἷν Cae et he , 
se t 3 y 
4 ἃ oe ΕΓ . εἰς δι td 
᾽ Ἢ 
> - ζ 4 ᾿ πὸ ᾿ 
+ 7 ¥ % ; 
Sith * εν ἐν, Salts | a, OS a τἂν 
rr > 7 
᾿ ὧν , » ῃ um 
sashes Ae) eer fox 743 
τυ Η a 8 Ὁ α΄ μὶ 4 
a 4 [ἢ ᾿ 
i ᾿ 5 io ᾿ πω. g 
ὦ ον , = 
tJ 2 ᾿ ua “9 - ‘ ; -* , 
᾿ ; ; ᾿ ᾿ ΠΕ λυ νν τ" 
dips: ΣΝ eee rit τ γον 
τ Ρ ᾿ - ᾿ > J 7 by ἀφ. ᾿- 
: ᾿ ᾽ ᾿ “Ὁ 
SELEY 7} Aes 8 ΤΌΣ Ὁ & ΘᾺ fos ὃ ᾿ 
Ξ- -. Ὑ 5 εὖ * ῥ: . i — Ψὶ 
a ae J { 7 r + » b i 
᾿ FS + ' " &, 
i f =k 
P os δι 
τ : Ν | Mees ing » 
: - ΝΣ ἀρ, kA My 
Ὡς yale : ss ¥ 
4 a _ ¥ " ἔ ΓΝ 
b ¢ Ὦ - = J 
« , "ΩΝ - * 7 
. ᾿ “ 
i é τι he "ss 
- - 
ae, - = - = ΕΝ = am - 
My : 
Ξ ? 
~ ΡῪ - > 
rf sie f ? = 2 αν =i, 
: = 77 2 
= ᾿ * 
cae) ‘ - 
om rok 
= ae ΟῚ 4 % a | : 4 
~~ - - - = 
£ > hm bu αὶ * 4 pe) 4 jee aoe. 
- = τς, 7 
ond 
4 Sy 5 Sea € 


LA 


Ἂ 
ty 


ANMAMAMAAMAAANMAABRRRHRBRBRWWWWWWNNHNNNNNND HHH HH HHH 


I. INDEX TO QUOTATIONS. 


MATTHEW 

PAGE 

5 109 
6 . 146 
ieee 109, 145 
ἘΠ Ὁ . 144 
τϑις 6] 
τοῦς: . 408 
20 120, 144 
Zafar 9 30, 447 
24 . . 453 
34: . 424 
Deine 29 
3 7] 
6 00, 108 
10 129, 445 
II LOG 
13. . 453 
20 f 453 
2 149 
4 431 
5 : WAT, 
II 10, 458 
12 10, 434 
14-. . 302 
17:- δ0, 458 
ὅλ 5 Ue 
ἴδ το : . 202 
. 13 112, 149, 158, 191 
ΠῚ ὁ - 3929: 459 
16. . . 424 
I7-. 455 
27" 405 
2s : . 454 
lige 132, 402, 408 
eG feito ὃ 90, 447 
18-. - 433 
20 - =) 65 
BP 152, 460 
32... 99, 236, 315 
BY ὁ 5 - 450 
35 f. : . 464 
36. ο - 129 
χες 3 154 
39 . : 450 
40: « q 424 


=e ee 


(a) New TESTAMENT. 


MatrHEw—continued 


Stas a oe aul 
12s . 448 
Ι9. ὃ . 432 
20. oad 
DR 6 5 on 422 
Dime : 206 


PAGE 
Be 4 : ο 67 
5. 48. : . 458 
On iit = . 471 
6: “τς: Ε . 422 
δ... : ; . 458 
6; ©) oc 5 . 419 
δείσας: 68, 272, 463 
(Oh, τ : ἘΠ 1 
θ. τ: Ξ τ 909 
6: 2:-- 5 . 419 
G26... ὃ . 16 
ὦ ). α : τῷ 175 
6: )ὃς- : - 195 
6. 30.. 10, 68, 202, 294 
Oundle : goles} 
“ΟΖ ες Ξ . 464 
Gs By ἢ“: : . 419 
TsO τι: : 75, 464 
Fa U6 83, 419, 421 
To Oe 5 600 
13. 2 . 419 
He WSS tc . 460 
χε 16. - : . 447 
ΖΘ: Ξ . 452 
ἢ: 22 208 
Th BE 6 . 409 
ΖΗ - : . 424 
7.25 se 190 
28. : . 426 
Soe iaas ὃ = Wz 
ὃ. 4 209 
8, 
ὃ. 
ὃ. 
8. 
8. 
8. 
S22 One . δον Ol 
Ss 27/6 - . 436 
8: XO) ac 3 . 452 
8:..22.- : 208 
ὍΣ - 5 206 
Cs BH 20 6 16 
9. 5 . 70, 206 

487 


MatrHEew—continued 


7,9 


10M. 
WE 
τ: 
oe 
ΤΟ - 
20: 
BOm. 


204, 432, 447, 


PAGE 
453 

- 426 
419, 441 
. 206 
129, 432 
. 453 

5 471 
192, 193 
. 469 

- 100 
424, 434 
204, 430 
147 

460 


5 ede 
29, 436 
245, 460 
430, 433 


4 - 950 
215, 432, 453 
: . 465 
10 

. 436 

128, 455 
440, 468 

128 

5 199 

458, 475 
106, 126, 189 


488 
MatTrHew—continued 
PAGE 
1233 421 
12. 36 Ξ . 424 
2 90 Β 403 
12. 39 . 468 
12. 42 146, 430 
12. 45 . 431 
12. 49 5 - 129 
TAG τ : .- 450 
13. ὃ - . 202 
15:15 00, 202 
17. ΤΆ 443, 444 
13: 15 74, 470 
ΤΩ tone . 437 
13. Lo:f. 205, 424 
13. 22 . 424 
13:23 205, 434 
1 20 Ξ . 450 
13. 28 421, 433, 452 
13. 30 = 4599 
152: 21 419 
13. 36 . 453 
13. 38 424, 44] 
13. 44 . 461 
13. 45 433 
13. 46 e452 
135,152 400, 433 
1353 426 
13. 56 467 
T3557, 464 
14. II - 430 
14. 26 461, 469 
14. 30 2 . 455 
14. 31 Ξ - 458 
δὲ - 443 
BS. 6-. 419 
D5. 15 . 424 
Th. 12 5 SB) 
15. 14 Ξ 71 
ΤΠ £5 - 453 
ES ΤΥ . 448 
15. 20 5, 193 
Th 23 - 195 
15527 345, 461 
15. 32 : 132 
16. 2 101, 405 
16.4 . 453, 468 
16. 6, II - 460 
16. 12 450, 460 
16. 19 7. 131 
16. 21 205, 455 
16, 22 . 455 
16. 24 421 
16. 28 464 
7 Ae 453 
17. 14 424 
τη. 15 ~ 405 
17. 19 ΕΝ 467 
17. 20 - 210 
17.27 . 369, 448 
18. 5 e729 


MarrHew—continued 


18. 
18. 
18. 


PAGE 

On. 442, 463 
We 461 
Ἐ2. [- 172 
.14 465 
. 16 467 
“τὸ 75 
. 20 434 
ἘΠΩῚ 422 
. 22 175 
. 23 3) 488 
. 24 432, 455 
. 28 5 298 
30 452 
Des - 426 
Ane 104, 160 
Ε 07, 402 

Ε . 469 

2 2 . 452 

4 - a ΔΩ 

8 103, 171, 424 
τὰς 5 3299 
Sie 454 
17 467 
21 438 
22 89 
23 408 
26 f. 458 
31 165 
ΕΙΣ ° 421 
33 83 
2 465 
3 458 
8 203 
x Oye - LOL 
II 112, 149 
I . 453 
19 108, 432 
22 . 247 
23 419 
25, 32 . 464 
33 191, 212, 433 
42 . 423, 462 
46 : 403 
eB 5 5 GBB 
16 464 
17 419 
22 453 
28 419 
30 . 410 
36 30, 442 
38 : . 442 

. 39 - sey 
ules + . 454 
2 . 458 
7 « 245 
13 - 465 
15 166, 441 
16, 18 . 464 
19 3 161 
20 ff; Fs 464 


INDEX TO QUOTATIONS. 


MatTrurw—continued 


PAGE 

25 126, 215 
725. ἘΣ 471, 479 
SAG) 121, 405 
5 SH Ξ - 472 
“538 . 433, 438 
36 : ον 471 
037 10, 133, 148 
aie . 424, 438 
ΞΟ Ὁ ξ . 447 
. 12 : . 424 
at . 144, 160 
20 - 419 
22 - 434 
26 5. 89 
29 . 451 
32 60, 264 
40 ‘ . 438 
41 00, 136, 356 
43 . 252 
45 : 339 
49 . 455 
51 ΜΠ 
ome 106, 206 
10 ἐκ Ll 
15 114 
18, 25 452 
29 424 
34 459 
Πρ ὦ 420 
12 - 445 
14 91, 420 
15 = is: 
17 421 
18 179 
22 455 
23 . 424 
31,33 . 464 
37 - 455 
42 - 468 
45 : - 421 
46 5 - 205 
53 76, 161, 422 
5 193, 469 
67 5 GB 
69 - 432 
72 . 469 
73 : τ 
74 205, 455, 469 
75 ore el! 
ΘᾺ. . 1189 
Ons . 419 
12 . 460 
24 461, 465 
28 5 ee 
33 105, 149 
34 Be 90 
38 : 438 
40 210 
43 . 469 
56 140 


INDEX TO QUOTATIONS. 


MatTTrHEw—continued 


sept εὐ aes aie CS cad 45 HOV AS 80) 109 Ὁ ΘΙ I et a ee at edn dee ae ee 


PAGE 
ὙΠ Ε 63, 400 
Oi 2 - 465 

I 439, 471 
4 ἐν 461 

5 453 

18 454 

19 400 

Mark 

2 400 
4 452 
Cp 6 . 104 
Gf = : 10, 434 
Sie Ε . 458 
On - 426 
Teles - 458 
12: - 452 
το - - 464 
TO}. . 423 
2 2 - 446 
Zoe: 446, 464 
24. 5 8} 
29, 30 - 446 
525 194, 209 
34 114, 189 
35 . 453 
38 - 2716 
40 . 206 
41 “199 
42 67, 242 
44 181, 420 
Asin . 446, 455 

Aes 324, 461, 470 
ἘΠῚ es - . 206 
6 Ξ . 452 
7 3 10 
-- Ε 432 
9 206 
14 453 
15 (427 
τὸ - - - 452 
19. - 419, 441 
22 . - 7206 
23. ‘389, 427, 455 
2Oigee Laer me ieee 
ik ἐν 223 
Aves - .-. 442 
5 = - 189, 325 
Ole 4 - 202 
12 « 446 
17 . . 44] 
19 ~ ° 145 
at 420 
22s 105 
2652 3am - 430 
52ὦ = = 206 
ἘΣ ῸΣ . 110, 455 
4 . 18, 426, 450, 451 
Same 176, 202, 439 


VIVID AAD AAD AD AAD AAD AAD ANAAANANNA AANA χὰ ψι σε σεσισισισισισυσισι ψισι ψισι σα ΒΘ ΦΦΘΘΡΡΡΡΕΡ 


MarKk—continued 
PAGE 


| 61,.68, 158, 


204, 444, 
60, 
436, 
231, 
319, 
420, 
436, 


12, 
76, 


446, 
92. 


211, 446, 
. 455, 
146, 464, 
- 1198; 
420, 439, 


194, 


205 
195 
470 
438 
439 
448 
468 
423 
162 
211 
419 
453 
206 
445 
464 
452 
446 
452 
195 
431 


Ὁ Ὁ Ὁ Ὁ Ὁ Ὁ Ὁ Ὁ Ὁ Ὁ Ὁ Ὁ Ὁ Ὁ Ὁ Ὁ Ὁ Ὁ Ὁ Ὁ Ὁ Ὁ Ὁ Ὁ Ὁ Ὁ 2 0 20 G0 90 90 00 00 0 0 0 0 90 90 00 00 20 HOH HIIIIIWAIVQYWYW} 


489 
MarK—continued 

PAGE 

9 100 
10 . 443 
ΤῚ 8ῦ, 153 
12. - 209, 419 
ΤΟ - 74, 448 
20. . 424 
Dal . 18, 453 
2° 420, 434, 452 
26 279, 349 
ey 5 345, 461 
50. - 5 199 
31 . 417] 
2 106, 129 
34, 35 - 89 
2 - » 447 
3 132, 221 
6 . 204 
7 194, 450 
ΤΟ - : . 420 
II 209, 430, 455 
12 - . 468 
ig}. 453 
τ π - 212 
ἘΠῚ 91, 400 
τοῖς . 194 
8 202 
19. 162 
Dit Ὁ 202 
2 436 
2c 189 
28. 454 
31 fe 455 
By 5 421 
SY 5 ΘΠ 
I 88, 464 
5 485, 452 
4 . 452 
5 453 
7 452 
8 468 
9 Ξ τ Ψ09 
2 - 99.110 396 
18 . 71, 257, 403, 409 
Ons : - 467 
21 164 
22 f 206 
20 . 446 
2.8) Ὁ 195, 467 
30 5 PAL 
23) - 209 
34 - 195 
Bib 5 458 
36 406, 430 
By) Ξ 5 wf 
2 442, 463 
3 - 106, 442, 448 
45, 47 - 442 
48 . - 206 
49. 132, 408 
50. 132 


490 
Mark—continued 
PAGE 
10. I 92, 453 
OP 2) 3 : - 195 
10; 6; <= 104, 160, 205 
10. 10 - ΜΓ 195 
10. 13 . 447 
10. 16 . 406 
0 Σ 17 - 432 
10. 22 420, 452 
10. 25 430, 441 
10. 28 - 455 
10. 30 157 
10. 32 452, 455 
10. 36 - 42) 
10. 38 25889 
10. 40 468 
IO. 41 - 455 
10. 43,44 - 458 
10. 47 - 455 
10. 49 - 104 
IO. 51 - 421 
EIS tu 148, 149 
ἘΠῚ 2 ἃ. - 420, 465 
BES 3) Ε . 458 
ERO ALS = (123 
bf eS ae ,. 16, 81; 375 
ἘΠΕ ΤᾺ, - . 459 
ἜΤ ΤῈ 2 - 455 
7F236 Ξ - 189 
11. 20 - 1055223 
BEL25 73, 204 
Ir. 28. . 419, 464 
II. 30 - 423 
II. 31 464 
σῶν , 212, 455 
2. 4: Ε . 995 
1 IS 245 
525675 430 
12. 10 - - 423 
12. 12 - 453 
12. 14 315, 419, 464 
12. 18 «- J95 
12. 23 = 8419 
12. 25 - 410 
12. 26 79, 123, 454 
12. 28 - 442 
12. 29 103, 454 
12. 32 - 208 
12. 34 271 
12. 35 Ε - 453 
12. 38 - 460 
12. 40 =) 60 
12. 41 .- 465 
12. 42 432, 452 
12. 43 - 165 
Ego2 190, 424, 438 
ΤΩ ue - 465 
2355 > re 455 
EZ19 - ΕΝ Od 
13. εἴ 424, 432 
13. 14 5 ,.. 222 


WNW ΝΙΝ Ὁ με ee 


ΔΙΑΒκ--οοπέϊπεα 

PAGE 

Pap f= : - 210 
S19 - 419, 435 
. 20 Ὁ . 434 
N25 : . | 45) 
S26 = : 447 
. 28 Ξ 60, 264 
- 30 : ei I 
- 34 ‘ ~ ade 
35 απο, 
Ἐπ . 420, 464 
2 91, 122, 379 
Ais Ξ = Ane 
ae =, WHOS 
SVE 445 
IO - 211 
12 - 42) 
15 = WEZE 
19 439, 455 
21 . 448 
25 445 
32 Ξ . 342 
33 - — SDD 

36 = il 
40 - 230, 452 
41 - 462 
42 ~ Ὁ 
44 Ξ . 190 
45 - - τ, 
47 - 432 
49 - 452, 467 
50 - 453 
54 f 192, 452 
3 Ε - 420 
65 - 455 
66 = 432 
69 - 455 
71 205, 455 
72 eee 
353 410 
6: : 319 
Fie 190 
Sis 455 
fo) 190 

2 421 

8 2 455 
oO 75 

I Ξ 67 

2 2 . 148 

3 202, 205, 407 

4 - 456 

5 : 421 
7 438, 447, 456 

> = 210 
34. Ξ, = 853 
36 - 109 
40 - 146, 452 
41 : Ξ 195 
43 - 452 
46 - 191,387 


INDEX TO QUOTATIONS. 


MarxK—continued 
PAGE 
TGs TON Ἔ ΦΉΣ 
πε 25... Ξ Ξ 439 
τοῖς - - 190 
16. 14 Ξ . 420 


LUKE 
E2 98 
i. : a: = tee 
ἘΠῚ gs s - 465 
1.8 426, 450, 465 
πξύτοις. Ξ . 452 
BS: Ξ - 103 
ας ον - 145,171 
Lp val ie - 450, 452 
ἘΦ -- 453 
ἘΞ 25 - - 426 
EN2A. 2 214, 245 
226 x 3 - 149 
τον = - © 193 
1237°< - - 434 
0301 2 - 453, 471 
ISAT = - 426 
a oa Ξ 443 
isle hg = - 482 
τὸ 57 ff. Ε ~ S417 
τὸ ἘΞ - - 466 
I. ΟΞ . 426 
EO? = = .- 315 
To F2.70 “5 - 466 
2.1 . 87,426,471, 474 
2: Ξ 60, 72 
2.5 - 105, 193 
rl C= . 426, 450 
2.5 Ἐ 3 - a SZ 
Pom hen - - 426 
2517 < : = 292 
ZOE) < : . 144 
Seer = = ~ 422 
ey τ Ξ - 104 
2» - - 450, 459 
2a25' 5 Ξ .- 406 
25:53 1 - - 466 
5 ae ~ = eh 
2: 37.2 = 60, 172 
2538) = - 432 
Ae fe tes : - 450 
ZA . 4338, 350 
2400 : - 426 
ZA Se - - 452 
2. ς = - 175 
3. ς - - 462 
ἜΗΝ ΤΣ Ξ - 455 
ἘᾺΝ Ξ - 100 
St ue) « - - 165 
Car ae - = ΣΟ 
S410 . 434, 458 
χη - 10, 434 
320 < - - 438 
So2E ak 427, 438, 450 


VVVVWYAADAAADAAAAADAAAA ADAM AVALAAALARAAAPLARARAAALDDRARADARADADRRROY 


INDEX TO QUOTATIONS. 


LuKE—continued 
PAGE 
22. é . 458 
22 Ol mare . 144 
Coyne 5 . 148 
TO. 450 
LOM: 149 
τ - 92 
ΠΩΣ Ε 67 
το - 57 
200 452 
Diu 455 
Ds . 440 
PG 123, 145 
26 433, 468 
PAG ᾿ 5 Ye 
ΧΟ Ὁ 191, 250, 453 
aii : . 452 
54: - 150 
36) 5 26 
Roms 452, 453 
BOF. . 453 
40. 194, 205 
Alte 6 . 448 
44. ὃ . 452 
Ts 426, 450, 452 
3 270, 432, 442 
5 : «208 
6 . 193, 206 
12 426, 433, 450 
Alas 2 > 204 
Ome 5 Ζ 295) 
Wg é 426, 433 
Zit ἃ 16, 455 
2.8) - . 453 
20, - 452 
34. . 44] 
38. 188, 224 
BOI : . 442 
τ ς 175, 269, 427 
4 : . 440 
6 . 427 
10) . 189 
II 2 E2khe 
12 427, 452 
TON . 145 
UG ic 158 
22a 67 
Boe 62, 98 
6᾽- . 458 
Some - .09 
BO 5 ἘΜ 
Ἄν. 103, 419 
44. . 447 
47... ὃ . 424 
48 101, 193, 250, 403 
ΠΈΡΑΣ - el: 
Fh. 421 
II 426 
12 60 
ts 455 
Pio 432 


LuKE—continued 


PAGE 
Ti 22: : . 453 
Tho OS} 6c 5 . 464 
7. 24. % 5 . 455 
ΑΝ τὸ . . 466 
7.28 . : . 459 
EES AN te 5 . 433 
ΘΠ Ὁ : ᾿ 46] 
Fe 38- ᾿ Ὁ δα 
5 42:1. : GS 
Vim AA - : - 127 
TE AG. : - 206 
20] - : . 455 
ἘΞ τοῦ τ : . 463 
Sa Es Ξ . 426 
δ᾽ .- 3 . 448 
Sot ac : . 450 
SuGree : . 375 
δ τὸ- 5 . 204 
ΘΕῚΣ" 5 . 434 
Sa eee . 424, 453 
$2 15) - : . 424 
8: τὸ. - : . 202 
Shy ἘΠ 436 
8. 22 426 
8. 23 . 994 
822 . 206 
8. 2 . 436 
8. 29 121, 386 
8x30). : eG 
8. 38 90, 195, 199 
&. 40 450, 452 
8. 42 60, 450 
8. 43 oe 117]! 
8. 48 . 463 
8. 53 6 195 
9. 3 100 
9. 6 420 
ὍΣ ᾽ς ὁ . 438 
Cyl so malig, cusses 
Cy 103% 6 Ε « Oe 
9. 17 5 Url 
9. 18 426, 450 
ὍΣ 23 49} 
9. 2ὃ . 426 
9. 29 . 450 
ὍΣ 232 + 901] 
Ὁ- 33 426, 450 
9. 34 103, 450 
g. 36 221, 450 
9. 37 . 426 
9. 38 OO 
9. 42 193, 403 
Os 47a. : - 919 
OL 5L - 259, 426, 450 
Qf 52... : . 466 
@) G3 - δ . 452 
ὍΣ 554. - . 421, 450 
Os BY 6 5 - 422 
Ὅε σδς ᾿ . 947 


HO} je & . 439, 466 


VOL. II. PART III.—33 


oe 9h 83 419. Το] 


4 F : 
. £5 450, 451, 460 


491 


LuKE—continued 
PAGE 


ges 5 . 135 
6. 4 . 441 


oe : . 429 
13. 5 2. 00) 
5 LL 5 . 405 
Dai . 432, 465 


meat 5 τ 47 
25 : = By 


ΕΣ é 7 181 
. 28 5 a GPAL 
. 30 5 . 24] 


. 34 fs LOS 
35 - 108, 450 
27 : . 466 
. 38 5 . 450 
. 39 - 145, 319 


241 : .« 435 
. 42 6 . 145 
Spice . 426, 450 


SS} 0 5 . 918 


73 
83 


. 14 . 426, 452 
- 17 Ξ . 438 
PLO ς . 480 
22 : {190 
. 26 : τ 4381 
2.27 110, 426, 450 
729 . 455, 468 
aS 146, 165, 430 


37 450 
39 3 . 471 
. 41 . 471, 479 

2 : 73, 256 


. 46 . 2 483 
Aas é ς ail 
= : or 
5 52 < 2 8455 
Ἐς . 455, 460 
Le Dis F . 436 
ΠΡΟΣ . 245, 460 
5 1) c . 424 
paler : . 447 
5 172 ‘ . 432 
12 ς ᾿ 450 


365 
447 


2/3 6 a 05 
e255 ξ Ὁ 175 
. 28 10, 68, 202, 


228, 294 


32 . . 346 
33 A 108 
36 5 5 ep} 
39 - 5 “δ 

2 ὃ . 999 
45 . 455 
46 Ξ: . 472 


492 


INDEX TO QUOTATIONS. 


LuKE—continued 


PAGE 
12. 48 424, 447 
12. 49 wt ATZ 
eS IG 103 
12. 58 191 
12. 59 , 210 
13. 2 . 488, 467 
13. 4 171, 424, 438, 467 
T3527. - . - 447 
ΤᾺ: ΤῸ - 452 
13,011 171, 172, 452 
r3. 13 . 190, 252 
18: ΤΊ » 100 
13. 16 171, 172, 447 
13. 18 Ξ a) Ἐ} Ὁ 
13. 19 462 
13.22 . 453 
ΠΗ 28 179, 455 
13. 26 . 455 
13. 28 218, 251, 448 
τῆ, 34 10, 130, 133, 

148, 245 
τὴς ἄς 426, 450, 452 
TA. 3) = . 454 
τῆ ἘΠῚ 422 
14... . 455 
14. 10 70, 210, 453 
15: 12 135 
14. 13 72 
14. 16 Ὁ» 123 
14. 17 "69, 70, 200 
14. 18 28, 455, 461 
14. 21 72, 
14. 30 455 
14. 32 15, 339 
14. 34 132, 402 
14. 35 447 
ΑΝ Gee 452 
15. Abit. 172 
1510). 135 
15 - . 442 
rh. 14 123, 455 
TS ΠΝ 433 
15. 16 63 
τς. 17 206 
15. 18, 20 . 453 
15. 24 . 455 
15. 25 452, 453 
15. 30 5 ead: 
TO: 2) - . 206 
16: 4 - τ 1.83 
τὸν 8 - 440, 441 
6 ὉΥ- 440, 447 
16. 13 = 419 
16. 16 113, 419 
16. 20 190, 192 
16. 22 427 
16. 26 464 
16. 29 - 146 
τ: 2 102, 193, 442 
τ. ἃ : 08 


LuKE—continued 


PAGE 
Var 70 
1 426, 450 
14 67, 426, ee 
17 Ξ 

ΤΟ : pee 
22 98, 100, pee 
23 
24 Ἐπ 
27 410 
Sie 75 
6. 440 
10 438 
12 994 
14 467 
29 67 
33 447 
35 426, 450 
37 ἀπ 150 
30 Ἐν 195 
41 - 421 
3 - 461 
4 ais 
5 5 ZAkl 
8 19. 171 
II 445 
12- δ. 70, 200 

.14 5 eh!) 
TS) 190, 211, 420, 450 
20 . 426 

2 60, 148 f., 152, 

369 
31 458 
36 203 
37 455 

43 421 
44 423 
45 455 
47 . 452 
48 . 192, 206 

Slats Ξ . 426 

sere ‘ . 419 

5 Oar . 212, 455 
Tie . 445 
17 . 423 
19 ney 
21 . 464 
22 > 419 
26 » 465 
28 . 204 
31 - 208 
34 383, 441 
35 410 
36 292, 383, 441 

2 19 
43 204 
46 é . 460 

Pllc a . 439 

Se3inc - 106]: 165 

τὸς . 424, 438 


ΜΕΝ 110, 364, 369 


LuKE—continued 
PAGE 
16 . 433 
23 slo 
25 126, 143 
26 2 7461 
34 230, 379 
. 35 466 
37 60, 152, 452 
. 38 : οὐ 07 
meer - . 145 
Ena te ᾿ Beall 
"Ὁ - : . 42] 
12 121 
15 443 
23 455 
26 . 458 
. 30 60, 207, 238 
31 5 405 
32 259, 420 
35 ΙΗ 
42 70, 197 
44 4915 
- 45 461 
ΙΝ 409 
= OL 11 
ΤΌΣ 453 
Dare 455 
Ain 34] 
ἊΝ 89, 454 
5 tes : 452 
ota a2 La 408, 458 
12 180, 348 
14 111, 341 
15 - 459 
22 τ peal 
31 114, 447 
35 . 194 
39 181 
40 198 
42 464 
43 92 
44 421 
48 194 
50, 52 424 
53 452 
ta 160 
3 100 
4 133, 426, 450 
- . 454 
10 οὐ 194 
12 : 453 
13 424, 452 
14 - 432 
15 420, 450 
17 - 432 
19 150, 465 
21 ay 108 
22 213, 358 
26 . 422 
27 noo 
30 426, 450 


INDEX TO QUOTATIONS. 


LuxkE—continued 
PAGE 
24. 31 5 . 461 
24. 32 83, 432, 452, 
472 
24. 33 453 
24. 39 121 
24. 41 461 
24. 42 . 343 
24. 47 450, 454 
24. ΕἸ 426, 450 

JOHN 

I 467 
Ade 5 . 436 
Ὁ. Ξ . 469 
San. E . 436 
QO. Ὁ Ὁ . 452 
TOLL oe . 469 
12 208, 424, 463 
13} ¢ . 436 
TAs 72, 162 
Tins . 474 
16. 436 
18 . 424 
2. 435 
230. 452 
20. - 474 
329 . 421 
2. - - 429 
353." 424, 435 
ΠΡΟ: ΜΠ 
Boi 152 
41. 269 
49. 431 
50, 51 161 
Bees : . 44] 
4. ϑ 5 te: 
Gr c : 452 
9: - 0 oa 
10 = GL 
II 105, 463 
13 : .: 422 
20 172,.191,.422 
22). - \- AT 
2B. . 463 
24. 5 . 180 
Samo =» = L955199 
10 . 422 
ἘΠῚ - . 469 
13,14 . 441 
Out 139, 463 
LS. 4 2463 
TO). 469 
Dig 193 
23 - 452 
26 - 424 
29. - . 444 
SB. - 424, 469 
36. - . .463 
Cee A - 431 


PWWWWWWWWWWWWWHNNNNNYHNN DN μη τ μμμμμμμμμμμμ μι μ 


ὰ ὦ τὰ AAADAADAADADAANADAADAAAAAAAAAAADARUMNAAAAAAAAAAAAARARRRRRRRRRERES 


JoHN—continued 


PAGE 
7 90 
9 49 
TO < 90 
We 6 74 
1 - 5 - 208 
20, 21 . 469 
23 365, 469 
31 - 10 
35 422, 469 
39) . . 463 
41 > [01 
46 378, 380 
49 210, 380 
2 5 (0 
5 172 
7 436 
9 190 
10 Ε . 422 
II 73, 139, 424 
LOM 5 . 424 
20 73, 202 
ὅλας . 469 
DEP Ὁ 431, 441 
28: - - 469 
29. . 431 
35. : . 225 
20m. 113, 161, 424 
37, 38 . 424 
30. Ὁ . 469 
40, 43 . 469 
44 429, 469 
2 189, 195 
3 5 Be 
13 oc 5 lel 
13. 5. 
το - 5 
25} Ὁ SL 
27) 9 - 441 
GIO) - - 463 
BOme . 436 
2: 9 05 
23 6 - 463 
Ch . .- 49 
39 424, 434, 437 
40 . . 468 
45 . . 459 
46 .. 424 
50 436 
Gs} - 44] 
54. 5 PAY 
δ o + 9422 
62° . 441 
63 . 474 
67 ΤῈ 
68 . 431 
70 - 469 
71 145, 171 
3 210, 476 
4 . 469 
5 463 


Ὁ Ὁ Ὁ Ὁ Ὁ Ὁ Ὁ Ὁ Ὁ Ὁ Ὁ Ὁ Ὁ Ὁ Ὁ Ὁ Ὁ GH Go 0 Go 00 00 90 90 0 0 00 DHIVYIVYIVVWYYVYYYYVyY 


493 
JoHN—continued | 
PAGE 
ΞΡΣ 210 
EF. 369 
18: 424 
10 - 469 
Zaire 139 
5 BO) c : . 469 
Neg ιν 165, 194, 463 
YEE} Ὁ : . 422 
Ba SOM Pe 00 
37 196, 474, 475, 476 
38 . 463, 474, 475 
39 . - 463 
41 194 
48 463 
51 211 
52 86 
4 193 
9. 12 454 
20. 469 
22 194 
26 - 424 
28 - 441 
30 - 463 
44 100 
45 436 
ἘΦ 469 
53) - 208 
fey 6 470, 474, 475 
BF Ὁ 221, 422 
Be . 48] 
7 ff. 32 
Sia 469 
10 194 
13. 43] 
TON . 194 
Tiare 83, 436 
13... - 491 
22 190 
25 474 
26) - 3 5. G8! 
30 83, 449, 469 
2. Ὁ : Seth 
34. . . 422 
a6 c . 441, 463 
BO: 435, 436, 463 
Alors - » 194 
ἘΠῚ: 424 
eee Be 71) 
23 146, 190 
24 194, 400 
25 .. 424 
29 74, 437 
40 . 452 
42 : . 463 
τς - 144, 452 
Bo : - 422 
9-. 171 
20 145 
25 463 
26 434, 463 


494 


Il. 
Il. 
II. 
Il. 
II. 
11. 
203 er 
11. 
II. 
11. 
II. 
ΣΙ. 
II. 
Il. 
II. 


JoHnN—continued 


PAGE 

28 211 
30 190 
32 145 
33 476 
36 194 
37 5 se! 
38 201, 476 
39 . 337 
44 72, 190 
45 . 463 
48 422, 463 
52 . 473 
54 50 
56 ey 194 
57 190, 211 
yc 5 39 
6 272 
II . 463 
23 . 441, 470 
34 423, 441, 469 
36, 37 . 463 
40 470 
41 409 
42, 44 = 463 
46 434, 463 
48, 49 . 424 
τ 5 470 
ΤῊΣ. 32 
PASC 211 
Aas 5 > Ye 
See . 368, 455 
yoke - oe 
10 248 
23 452 
24 . 204 
26 345, 435 
29 211, 272 
31 . 441 
I 463 
2 δ - 470 
ας Β 60, 247 
TP 3 oe 
12 424, 463 
13 . 424 
16 422, 436 
17 00 
21 424 
22 181 
26 . 424 
31 211, 470 
2. Ἢ .. 424 
Ye 448 
16 211 
22, 24. 194 
Bo 470 
5 469 
Ore 463 
20 . 462 
24 247, 421 
25 - 469 


"μμμιμμιμιμιμιμι μι μ 
ἘΝ κού αἱ ἐγόν 2. τ πα 


JoHN—continued 


PAGE 
. 30 - . 463 
32 . 204, 470 
.2 218,424, 437, 441 
Ber ie Ξ Ὁ 
Ome : 221 
7 215, 221 
8 - wld 
II 437, 469 
12 τὸ 437 
20 . 463 
23 ‘ τ, CYR) 
24 60, 424, 437 
. 26 Ε - 19 
I 121, 149 
a 4598 
18 424, 452 
25, 30 . 452 
32 . 444 
33 32 
34 181 
37 60 
39 421 
Sie 202 
13 409 
17 148 
23 67, 170 
31 189, 226 
38 a BY 
39 - 106 
vier . 439 
2 448, 474 
Fe 5 BI 
12 438 
17 476 
18 474 
19 439 
23 206 
25 129 
29 469 
30 465 
6 461 
8 141 
II 469 
15 165 
18 203 
19 . 444 
25 216, 219 
Acts 
ὙΠ": 9, 455 
δ: - 195 
TOM 133, 452 
2) ἐν ΠΣ 
13, 14 452 
ic C . 473 
18 . 49, 246, 323, 440 
TO} - 5 49, 153 
20 : 157 
22 454 


INDEX TO QUOTATIONS. 


Ἔα τ τὰ τὰ Ὡς σὺ οι οι συ ρισισι σι ισισι σι σι σι ΒΘ ΕΘ ΒΕ βῳφῳφῳφῳφωφφο PP RYE RPHSPEYPPHLHP 
OV 


Acts—continued 


PAGE 

I 426, 450, 473 
2 5 . 452 
4 455 
bias é . 452 
op ὃ 98, 100 
je tee 5 - ‘asl 
7a 443 
22. 462 
Diss 190 
26. 191 
50: 443 
42 452 
46 - 195 
Agee a les che 
Be” os . 202, 447 
δι - : - 100 
i 112 
IO . 126 
Me ς 146 
ine} . 466 
LON. 465, 473 
19 - οἴ 220 
20. 466 
BSc δ al 
20): 450, 451 
Awe - oe its 
Rc . 426, 427 
@) Ὁ : . 260 
ΓΕ ἢ 112 
τ - 443 
19. 453 
22. 190 
25 > . 474 
30. 450, 451 
83 3 202 
asc 202, 206 
36: . 44] 
2.3 408 
- 458 

7 420 
121. 146 
τ. 206 
Τοῖς 102 
Wipe 126 
28 . 443 
29. . 453 
36. 96, 462 
39 - τ 9 
Alas 466 
I 158 
. 162 

5 129, 162 
7 . 473 
II . 440 
2 - 4102 
7 Eo 
10 - 465 
II 151, 192 
12 22 
19. 309 


ee OO ENO ΧΟ ep Oe 00 Θ᾽ 22105) Go 00109 G0) 00) 6050) 69:08: Co 1S YY ST Ten eet ae oa 


INDEX TO QUOTATIONS. 


Acts—continued 


20. 
2I. 


172 
214, 444 
123 
151 
429 
384 
146 
466 
191 
463 
44] 
453 
- 450 
194, 474 

- 452 
194, 202 
. 465 


189, 266, 
146, 


83, 302, 


301, 
157, 


497, 
147, 427 


108, 453 


210, 453 


100, 


Acts—continued 


PAGE 

Pap ce) . 210, 461 
120 72, 163, 427 
. 28 . 123, 474 
Suse ὁ 445 
odhig < 5 {1710 
= Bite) 3 . - 452 
FP. : . 210 
. 10 83, 158, 432 
: 1Δ4 : . 461 
. 20 . . τ 452 
ἜΖΗ. Ξ 87, 191 
ἄπο ς 88, 91, 474 
aoa 5 » + 128 
Sm. : 83, 165 
= 10 δ . 441 
512 : - 255 
. 18 58, 109, 390 
. 19 5 = 991] 
. 20 : 5 ily? 
22 : . : 462 
24 5 . 466 
25 . . 434 
21 : αὐ 10] 
47 . 450, 462 
. 48 : τ 161 
ΕΙΣ 6 - AZT 
ΕΣ . 388, 452 
ΤΣ .ῦ 199] 142 
say 4 = 102 
3 UG) : 91, 106 
. 19 . 194, 240 
5 i ‘ . 400 
23 - 190, 463 
27 : . 466 
ab . 148, 466 
59s Ὁ . 474 
.Ὁ - 5 oe dil 
ly, : . 435 
. 20 : a) 98. 
. 28 .69, 70, 165 
29 : - 109 
"532 5 - 259 
ee - 195, 202 
5 - 5 - - 192 
BY : . 194 
Ole : - 5129 
BGC) - 452 
4 pitt 100, 151, 278 
12 ‘ - 452 
16 100, 427 
18 - . 432 
ἘΣ - 102 
26 121, 122, 190 
36 : . τ 22] 
ΠΩ 5 - 453 
- 39 - 195, 453 
ΘΓ 57, 194, 389 
OL. - 129,194 
5 Fc : » 465 
Fos : - 165 


495 


Acts—continued 


PAGE 

. 18 : 73, 275 
ΣΟ ᾿ 15} 
21 : eo 
522 5 « Ld] 
. 26 5 . 466 
5 27 196, 211, 216 
. 28 . 9, 73, 244 
hat . 386, 452 
12 δ . 9995 
ΤΊ, 6 . 189 
ἐὺ 5 - 39 
18 : 5 08 
. 26 . . 212, 45 
Ε 2 : . - 160 
. 28 - - 205; 301 
ao cae 121, 427, 450 
BAL 6 463 
oho 171, 172 
12 se 72 
13 . 365 
14 . 194 
25 - 10 
27 111, 462 
28 . 162 
29, 31 124, 369 
31 . 124 
32 67, 161 
a7. . - 125 
38 57 
40 341 
7, 439 
ὃ 328 
9 . 286 
14 72, 79 
15 113, 362 
16 148, 392 
18 32208 
23 12157122 
111 

ers Oils 101, 121, 427 
By: : . 452 
ahs ξ . 4271 
Vane : 129 
10 . lo 
ἢ 22] 
: : « 4 
ἜΣ 91, 188, 195, 265 
- 35 ᾿ . 427 
5.3). 4 C . 144 
τ... Ξ - 427 
. 10 : . 453 
5 abi 0 . 461 
13 : . 432 
ΙΧ : - 427 
19, 20 452 
23 3 & UPD 
25 - . 407 
BS hirs 5 . 206 
pO)ac ᾿ . 389 
ΤΣ Ν 89, 90 


INDEX TO QUOTATIONS. 


Acts—ctontinued 

PAGE 

23553 - 161 
23. 14 . 443 
23. 16 - 125 
23. 20 - 212 
OE as 89, 90, 161 
2a5 ΖΑ͂ Lakes 
23. 30 . 219 
24. 2 354, 455 
PAS ς - 89 
24.9 - = - 212 
24. 11 161, 172 
24. 16 - 463 
24. 27 19, 132 
25). 157 
251 ν- 161 
25.9. 132, 453 
25. 14 161 
25256 . 466 
26.4 - - wel 
205 7 216, 219 
26. 19 - 157 
26. 21 67 
26. 26 111 
26. 29 192 
26. 31 454 
26. 32 190 
Στ πες Ὁ 202 
κεῖ τ ΩΝ 106, 199 
ZEN: « - 427 
Ze α 101, 128 
27257 - 205 
27. 19 193 
27. 20 : = 96 
2727 71, 175, 324 
27. 29 192, 324 
27. 33 111, 175 
27. 34 111 
27. 35 455 
27. 39 108, 309 
27. 40 Ξ - 321 
27.41 9, 118, 243, 294 
285 τὶ - . 959 
252s = . 194 
28.6 60, 106, 194, 197, 
207, 254, 384 

28.7 . - - 27 
28.8 . 125, 342, 427 
28. II 88, 271 
28. 15 = 109 
28. 16 ae 7! 
28. 17 . 427 
28. 18 - 188 
28. 20 67, 319 
28. 26 oe + BS 

Romans 

Les 2 75 
I. 21 191, 395, 463 
Tare : - 402 


RomMaNs—continued 
PAGE 
1.24. . 463 
¥. 26) . . 440 
27. 104, 309 
% Jor. - -2ia 
ZG, - . 440 
2.15 - 206 
22. - 450 
22. . 463 
3. 4 - 450 
3.9 322, 385 
3 11 60, 202, 205 
35: 3 : bo: | 
cele 108, 195 
cet: - - 465 
3. 20 . 434 
4. 3 463 
4. 5 139 
“ες 190 
4.37 « 465 
4.359.. 58 
op a9 5 op 
5. ie - Ξ . 463 
ete 190, 196, 198 
er - : - 463 
6.6 . 440 
7-9 194 
7. 10 60 
7-14. 378 
7 ae ae 61 
ἈΡΑ͂Ι ς 445 
8. 20 98 
“ τὶ. 440 
8.26 . 200 
a ae 219 
C.F > « - . 463 
re 23, 441, 463 
9. 16 . - . 197 
9.18 . 195 
9.29. 190 
10. 14 - 463 
10. 16 Β . σῇ 
te). : - i23 
ay ΥῊῈ -. 152,173 
FES. . 450 
II. 10 124, 450 
II. 25 -~ 3 
11. 35 422 
ps ae 378 
12. 20 196, 440 
12 25 - 199 
Ee - >. eee 
13. 12 ς 212 
14. 4. . 390 
14.8. Β 7 
14. 15 - 205 
14. 19 - 14 
14. 21 89, 90, 140 
14. 22 - - .180 
5 a "ΗΒ Ξ ΑΒΕ 
15. 13 - . 450, 451 


RomaNs—continued 

PAGE 
15.27 378 
τευ. 902 
16.2. 60 
16.0%. 144, 446 
τὸν. 17, 221 
16. 11 139 
16. 12 446 
16. 15 350 
16. 21 88 
16. 22 τ . 485 
16. 26 379 


τ᾿ Ξ - 

Tt. 20%. 365, 402 
Ἐς 22>. = 1. 
Fe iy ΤῈ 444 
ΖΕ ὅς 78, 143 
Ze δ: 459 
2 1G... 187 
3- I, 3,4 378 
4. ὅν 241 
3.9 941 
3. 14 60, 191 
4.1 . 433 
4. 6 70, 75, 121, 200, 463 
4.9 . . 9680 
4. 10 307 
4. 11 72 
4. 21 23 
ae, 112 
Coe fie 60 
epic ae 181 
6. 2 60, 343 
6. 3 > eae 
6.4 Σιων. -- 
6.7 . 9854 
6.9 104, 377 
6. τι . 9391 
7- 3» 5 - 203 
7.9 . 309 
y Be 60, 203, 325 
ype oe . 429 
7 et oe 463 
7. 2D 419 
“ΣΤ 49, 165 
γη.26. 433 
y ee 6 oe 156 
7. 30° . 352 
7. 38 409 
8.10. 342 
8.11. 206 
5. τῷ - 140 
QE. = 193 
9.4 90 
6.85 - 395 
ene yo 78 
9. 53.2 399 
9.18 . 200 


INDEX TO QUOTATIONS. 


1 CorintTHIANS—contd. 
PAGE 

9: 21 60, 243, 307 
ΘΖ. : Ὁ 70 
TOWAS 208 
TOS ers 90 
TOMO 3 .« 206 
10. 10 71, 206, 365 
10. 20 - . 448 
10. 22 eG 
10, 29 : : 151 
ig (6) ¢ 199, 200, 250 
Hits, 18; 20 5 lz: 
ΤΡ 1 195, 450 
11 22 . . 226 
11:23 . 206 
11. 28 433 
II. 34 195 
12. 31 - 202 
Tete 111, 205 
Ἐ2» sis 74, 111, 219 
13: 4. . 400 
ΤΩΝ. 60, 204, 409 
14.8- LOS 
14. 18 473 
14. 19 442 
14. 20 405 
14. 23 473 
14. 38 70 
BX, (ro 315 
HS. te 60 
15.9. 166 
U5 7, 2) Lb 
15. 19 se Ὁ 
15. 24 . 204 
15. 27 eo 
15. 33 9, 61, 63 
15. 49 5 74 
τ: [2 . 330 
ΤΡ ΒΖ ... 120 
1π5ὲ δ 22,119 
Sew 5 0, 7: . 126 
16, 2 191, 200, 203, 
238, 439 

TOWN - . 204 
16. 6, 7 467 
16. 11 112 
16 12, 10 . 446 
16. 22 154, 204 

2 CORINTHIANS 

τοῖς 5 199 
1 2 Ὁ “975 
᾿ς . 295 
HE ΠΟ Ὁ “110 
2.1: - 165 
2: OF" - πὸ 72 
2. ΤᾺΪ α . 400 
2.1 - . 465 
Soak 205, 379 
32 . 295 


2 CorINTHIANS—contd. 


OCOOWAAAANMUA DRAW 


Re GEE oe See ee θὲ δον Θὲ. apes Se Ge ge Be 


PAGE 
6 . » 245 
10 : 67 
16 61, 321 
2 205, 212 
6 - 105 
Sim « 5 GD) 
TO: . . 439 
SH 5 . 467 
LOI. . 452 
2 . 440 
4 205) 
Gh - 287 
τῆν: . 900 
18 . . 468 
12, 72 100 
2). “ὃς - 1290 
Gc 191, 366 
5 ΠΣ . 390 
Tare . 466 
Dee. 2 102 
4 71, 443 
10 5 ΜΠ} 
ol. lll 
On. 467 
23 326 
24 5 leis} 
25 193, 269 
27 . 125 
BD ¢ 172 
Gigs . 463 
Ole 165, 463 
13 .~ 396 
15 165 
19 465 
Sh: 390 
GALATIANS 
18. 387, 467 
LO}. . 468 
22 2.2 452 
lems 172 
4 dom LO 
5 191, 467 
1 . 466 
7) 327 
TAs - 100 
16. 463, 468 
7? ¢ >. 60 
10 450 
9, 172 
TOs 113 
23 104 
4 . 459 
7 ; 75, 196 
18 69, 200, 450, 467 
19): : a 7119 
70. - 407 
22s 438 
DG) 157 


ΟΥ̓ΘΕῚΣ 


PRRPREDPLO OOOO NPS Yanan oe 


EXON ON TONER ON Gon i 


NNN NN HHH μι μι 


497 


GALATIANS—continued 


PAGE 
28). 23, 441 
20 126, 131 
II 3 8 
12 74 
EPHESIANS 
es 444, 485 
= . 465 
6, 8,9 . 419 
14 . 440, 485 
LQ) - . 419 
22a . 485 
2: 3 . 44) 
ἤν . 419 
6 . 409 
8 . 200 
17 453, 485 
8 . 166 
Mite 419 
τς 193 
TS. < 310 
19, 20 419 
Ι,4 419 
ΟΣ . 166 
II . 485 
14 339, 408 
7) - 485 
22 440, 485 
20, - : . 434 
ἘᾺΝ 29... 203, 222, 
434, 444 
6, 8 44] 
14: - 210 
16 . 309 
26. 369 
27" 361 
290 - 911 
ἅς 121 
31) 463 
ἜΝ ς 311 
Ol. 205 
190 339 
ro) 347 
1@) ¢ 485 
20) « 369 
PHILIPPIANS 
12΄- 122, 403 
18. . 54 
16. - "990 
DP) Ie 408, 422 
20 « . "407 
720) . 403 
Tak - 73 
τὸ - 61 
22. 98 
26) 452 
Do) 173 


498 


PRERROOE HYD 
οο 


+ 


PABRRRWWWNHNNNNNNH HHH μμ 


σισισισισιιια fp Bw ww ὮὌ " 


ῬΗΙΠΙΡΡΙΑΝΒ---οοηϊά. 


PAGE 

30. - 400 
5 . 286 
6 - _ 126 
8 > 205 
10 396, 407 
21 . 440 
2 150 
287 

10 . . 390 
12 197, 405 
15 5 BBC 

CoLossIANS 

DP : nee UG: 
ἘΦ, - 193, 441 
TiS ers : mee) 
TS 5 5 909 
Oe. . 440, 465 
26. - ὃ 499 
920 - Ξ - 419 
ΠΩΣ τ ao 195:.22} 
7s 3 95 
Tare 108, 310, 440 
1 310, 400, 408 
TOM. ᾿ 91 
18 273, 290 
23 - 290 
6 23, 441 
9 . 310 
19 402 
5 309 
Ome : . 292 
12. : Us 
HEY 6 > 55278 
iti, Ἢ : 71, 118 


1 THESSALONIANS 


110 
275 
251 
467 

73 

419 
485 

60 

. 99 
22, 441 
: F > ates} 
cor Ξ 311 
δε: ‘ 5 PANT 
TOM. 106, 203, 257 
: 5 1 ATE 


HHT OHO Of CON 
wat 


2 THESSALONIANS 


110 
466 


iO " 


2 THESSALONIANS-—conid. 


PAGE 
ΦΧ 23, 441 
Di vik τς 20D 
P36) oS 467 
Py he 228 
DAO, 157 
πὸ ς 259 
3346; >. 209 
ἘΝῸ πο), ὁ . 467 
Φ ἜΑ͂Ν - 207, 396 
Zon. eer lit 

1 TimotHy 
1G Hye - . 200 
1. - 68, 274 
I.10 99, 104, 315, 365 
Ng LEE Ss Ξ - 366 
2. 8,9 : ΘΠ ΟΝ 
2 HID ὦ : seers, 
Ἐν.) ἢ 5 76, 379 
3: Shi. : 76, 157 
Zone : 47, 417 
Cal : a LOS 
δ: ἜΘ᾽ - = 5 -90. 
5. 1n3e : - 158 
S22: a 20350207 
Gabe. : 5» 90 Ὁ 
ραν ΠΣ - . 443 
6.18: - 91, 378, 388 
οἰ σον: ὃ τ (69 

2 ΤΙΜΟΤΗΥ 
Gots 19] 
1. 6 284 
ee). ὃ 5 GH) 
Hep lO Oe: 
ΠΤ. 18 164, 211 
2.4 - 389 
RiGh 3 3 5 eA 
2 Aw - ce 989 
Do τ Ὁ 5 3 wd 
Στοὺς : ἐς, 09 
Ζ: ΣῊΝ" Ξ 9.719) 
Spe ie ὃ 5 Ma! 
Shelksi nc : μῇ ρας ἐλ]! 
4: τοῖς ὃ <a 6d 
4:03) - ς πὸ, 106 
2. τ - : 85, 211 
4.16. Ἴ oe 2A 
Anat . - 193 

TITUS 

1.95 395 
τῷ : s Wey) 
Τ 12 . 9, 137, 158 
ips 105, 223 
2.2 =~ 46 
2533 278 


INDEX TO QUOTATIONS. 


Trrus—continued 


PAGE 
Perit ee i) 
ὮΣΕ 273, 274 
2.8 5 ISS, 
2110 - 408 
Ziel. 1s 157 
310 tee Wels ΠΟΙ 
3.710. . 379 
PHILEMON 

2 109 

9 86 

II 353 

12 Sales 

13 188, 467 

rs! . - 190 

201. 213 

HEBREWS 

ἘΠῚ τς 250 
lig 5: 298, 368 
ἘΠῚ . 463 
2:18 450 
3: Ἃ 49 
3. 5. 220 
Ἅ π1Ὸ ks 408 
2. 408 
3. 12 25, 450 
Se RUSH 7: 113, 439 
4. τος 25, 450 
Sh MOS 1 60 
ἧς On 216, 219 
4. 3, 5 - 468 
4.9 901, 409 
4. 12 83 
4.-π3᾿΄- 407, 407 
ΠΝ 461 
5a Obre 399 
Gone - 60 
ONGie 394 
6. iwc 444 
6. 10) 6 ; 7 + 139 
5 8 191. 9515 276 
TWAS Ere 3 - ' 24 
7 MONG 378, 379 
TP ozs: 
G5 PIE Me 371 
δι δὺς: 129 
8.6 262 
8. ὃ - . 422 
8. 10 429, 463 
Sarksh ἃ . 450 
Θ: 2 οὖς OO 
9. 3 . 443 
9. 4 . «124 
9. 5 oe 
9. 12 157, 213 
ΘΕῸ: 21 74 193 
10. I 82, 448 


INDEX TO QUOTATIONS. 


HEBREWS—continued 


PAGE 
ΤΟΣ 2 ᾿ ἘΠ 
10. 7 . 450 
10. 10 25, 463 
10. 16 . 429 
ΤΟΣ τ 72 
10. 21 129 
10322 192, 248 
10. 25 78 
10. 33 - - 407 
10. 34 5 US 
ΤΟΣ 7; 265, 270, 442 
ENS} ᾿ς Ξ : 1905 
τα το 278 
LRU 85 
ΠΤ 199 
11. 18 463 
ΤΎΡΟΣ 121 
its Γ23 78 
11 28 71 
iis ι 109 
II. 32 . 103 
Lied! 105, 286 
IT. 35 69, 83, 407 
ΤΙ 1. : 5 UR 
is ie 178, 282 
UDP - - . 305 
ΤΩΣ ἥν 113, 189 
175 ISG - 108 
125 12 . 440 
ΤΏ ΤΕ 258, 264 
12. 16 é 5 2 
12 221, 222, 310 
12. 18 Ξ 5 ae 
12. 19 125, 126 
12. 20 375, 399 
As Dre 911 
ΘΑ. 108 
Eon. 154 
13. 15 139 
13. 20 25 

JAMES 

LO 235, 407 
Uy 1G 5 mS all 
Testis : . 247 
Hig, TESS 60, 246 
ΤΟ ΠΟΘ ὁ a 90 
HG iy e o- Cb 
ἸΣΤΟΥ: . 222 
Te 2 πὸ- - 105 
122. 27, 235, 440 
Ι. 25 27, 440 
ie Ours 157, 290 
DAOMe re - . 845 
Dao is 180, 378 
Poe, 118) - - 156 
2. 20 158 
2. Pas. 27, 109 
3.2 446 


ψισισι σι β ββ β βεε β  φῳῳῴφῳ 


UmmummnPBPRWWWW dH YH ND DN DWH μι μι 


NNN HHH HR 


JAMES—continued 


PAGE 
3 U7 
5 391 
δ᾽. 59 
9 27 
10 27 
12 80, 377 
4 > . 222 
5 = . 407 
7 - 42} 
ὃ 234, 421 
9 "Ὁ 3206» 
TO) - -. 49] 
12} - . 400 
ww 5 8B) 
Gi 5 : 74 
ae - 395 
4 221, 393 
7 5 BiB! 
12 = 204: 
Ἐπ - 23, 444 

1 PETER 
ye 217 
St. . 463 
Τάς 20, 441 
07) ΠΥ AE 
18 . 157 
ΣῈ: 378 
5 211 
6 321 
II 378 
2 440 
ρ.- 435 
Zbl 452 
I 74 
4 160 
IO): 450 
15 299 
5}- 193 
Gh 09. 22 
19 5 Pv alos 
21 ΣῈ Ὁ 2005 350 
6M: : .. 199 
8: .“ .60, 89 
TO} = 396 
12 110 
2 PETER 

2 : oa 
Ἴ-: : - 165 
OM . 290 
Ie ce ; 5 Avy 
17. 5 1718 
LO}. 87, 379 
2OWs . 434 
3 . 400 
4 78, 396 
6 147 


Ln ll on Bl oe 


WWHWWwWNHKHNNHNNN 


ψισισισισιθ ΟΝ ΟΝ NHN HHH 


499 


2 PeTER—continued 


PAGE 
ΕΞ - 105 952 
12 : - . 420 
νὰ 219, 420 
TAS. 87, 441 
ΤΟΝ Ὁ - 440 
ἘΞ. te 3 “165 
22. τ Ὁ; “515 d+. GOS 
Sue is . 366, 420 
ΘΙ: ne ffs: 
10 242, 395 
12 =, 268 
ἘΣ - 959 
1 JoHN 
2 . 467 
One . 242 
ΤΟ" 190, 433 
Phi . 434 
DAs 424 
2 sare 157 
26): 197 
Zp - 208 
28. - 460 
10 441] 
POA. « 5 . 465 
Ay is . 126, 437 
Sake. . 462 
ΤΟ 2 463 
20) 74 
21 460 
2 JoHN 
2 429 
5 194 
Sas 3 . 162 
120 - 188, 204 
3 JOHN 
4 166 
6 465 
9 399 
JUDE 
ae 5 5 aly 
AMG. 132, 205, 209 
12 279, 358 
18} eg 162 
ΤΡ Ξ 192 
Ge 235 
16, 18 366 
225 22 196 
Dhar. 465 
REVELATION 
AGe - . 154 
Gua : 79, 429 
6h. . 429 


500 


INDEX TO QUOTATIONS. 


REVELATION—continued 


πα SOO SNM Gy Dy Qu ON) ONE aE REO COSCO OCI ON) C9) 9) C2 SANDTON BENS TENT τ το 


-IO:. 


II 


‘io, 120, 139, 441 


᾿ 34, 206, 230, 429 


168, 169, 173, 454 


PAGE 
, ΘΠ 
106, 128 


280 
- 131, 429 
120, 121, 359 


221 

215 

- 221 

197, 250, 424 
: . 106 
429 

151 

119 

. 467 

197, 250, 424 
. 280 

89, 202, 429 


Ξ . 424 
210, 421, 454 
. 157, 359 
67, 121, 170 

67, 157, 170 

67, 170 


- 159, 222 
120, 121, 170 


454 

170 

. 439 

75, 101, 424 


REVELATION—continued 


7. 


ΘΟ ΘΟ ΟΊ SONGS COL O9, COLOR CORT IOS 


PAGE 

GO: : . 435 
Tikes ᾿ . 190 
TA). 5 . 429 
τό: - . 434 
1 - ὃ =» τ OA 
ες s+ 218; 9. 
es : - 123 
II 123, 343, 462 
12 ς 60, 262 
13 . 278, 432 
4 - 29 
8 . 194 
II 205, 476 
13 . 432 
16 168, 173 
19 2 167 
20 75, 121 
I . 476 
2 347 
6 464 
7h 422 
Siac : . 347 
ΟΣ: 208, 347, 402 
τὰς ς . 454 
2 172 
7 467 
9 178 
10 264 
II 178 
12 210 
15 454 
17 221 
19 476 
Bi ξ 101 
oe : 104, 161 
MOE = 15%. 435,448 
.7 448, 449, 467, 484 
. 10 Se 27 
ap » . 100, 423 
13 104, 129, 161 
era . 435, 466 
S19) - . 467 
͵- - . 440 
De : 112 
Bae : . 476 
Beth. : - «407 
Bie ᾿ ἘΠ 
7 . 467 
8 . 435 
10 449, 450 
II - 476 
12 - 2450 
14 - 129 
m5 75, 430 
18 - 169 
I 160, 222 
2 . 429 
Sie 173, 429 
Gb. . 454 
9, 10 422 


REVELATION—continued 


PAGE 
13 o = le 
14 121, 441 
19 : al23 
20 123, 173 
2 67, 197 
Bhi. 429 
ΠΡΟ 470 
ale 91, 196, 265 
Dic 126 
Spe 90 
Or. 445 
16 149 
19 462 
ai 440 
cee 60 
6, 8 476 
One 435 
.14 467 
3. 127; 291, 381, ὩΣ 
ἈΝ" 
ἘΠ 2 - 155, 362 
ΙΖ : . 335 
125 ; - 432 
22 ᾿ . 434 
5 BS < 60, 262 
ae «+ 22423 
EVAN ts 5 oe LO 
shel : : ws LS 
3 . 54 
Mis 192, 230, 256, 321 
as . - 123, 436 
2 116) 3 . 443 
ral : . 123 
19 . 467 
I . 131 
4 . 429 
8 171, 435 
9 - 400 
II . 466 
6 221, 424 
8 349, 460 
10 . 124 
II 123, 407 
16 = lz} 
517 Σ 141 
. 18 67, 124, 307 
LG . ὁ L225 376 
.20 79, 90, 124, 362 
HoT j 67, 303 
27 434, 468 
ἘΠῚ: b(t) 
4 129, 205 
4: eo aoe 
LAG ον, . 264 
δος = #209 
ΘΕ: . 54 
παῖ 395, 400 
17 . 475 
18 173 
20 154 


INDEX TO QUOTATIONS. 


(Ὁ) OLtp TESTAMENT. 


501 


N.B.—The numbering of the chapters is according to the English Bible; 
where the LXX differs, the numbers are added in brackets. 
titles of Books. 


PAGE 

δι 0 Ὡς 17 
ἘΚ ΣΙ ὩΣ ane 122 
ἘΠ τ Ὡς τς 439 
ee 2..:2 ἃς 455 
21: ΤΥ 443 
pda ee Oe 438 
ees LO 444 
A ow. bee 175 
Be AOS νος 195 
Ὁ: 10 - 439 
Ge & 439 
5 Fe). 442 
». (Oe TI 461 
ΟΣ BBY c 124 
ess 11 437 
ela 23 469 
Oa . 434 
se 18..25 448, 449 
ee T+ 20 . 468 
τ ΖΟΡΙΙ 400 
eon 12 439 
sn 26:29 = 190 
275. 31 428, 444 
sen 27s 30 . 447 
ay | Py 06) 469 
9 29. 33 463 
oy 30.5.38: 41 123 
co karat 447 
"4153 420 
33 Als 19 435 
33 41:42 477 
99) 42038 426 
» 43-34 438 
» 24. 2S 469 
op A 21 438 
Ex τ. 12 442 
"11 154 
SLA: 442 
"Ὁ: 27 435 
ΤΙ Ὁ 435 
ap WAS ἢ 437 
ee ἘΠῚ 20 438 
FLOSS 3 124 
53 Ζ28. 20 362 
ie 35 477 
"20 [ἢ 410 
Lev. 2. 13 132 
Num. 14. 7 442 
ese 4. 22 447 
» “20. 10 152 
28. 13 (14) N27 
Deut. 15 21 990 
eG 447 
» 4.32 438 
A) Gio 22 439 


PAGE 

Deut. 8. 2 309 
aS 4! 447 
Pano, 10 309 
ἐν 28. 412 442 
eT Oy δ. 476 
» 28.64 206, 438 
Se 2053 448 
τ ΘΙ σῷ 440 
Jos. 2. 24 469 
ay 2. 10 442 
UGA 6S} 443 
ΤΟΣ 1S) 153 
Judg. 9. 34 171 
τ ΤΟΙ τς . 447 

1 Sam. (1 K.) 2. 13 425 
oe ΤΊ . 428 

7. 16 439 

oy 135 14 402 

» 14.4 438 

» 14.45 468 

oe E70 476 

Ὁ 1. τὸ . ΡΣ 

2 Sam. (2 K.) 11. 11 468 
a 46 438 

A 1H OR - 149 

» 19.21 448, 449 

I Καὶ (3 K.) 14.6. 198 
se Le 13 149 
Pela LO 433 

5 ΤῊΣ 12 940 
18: ΤῸ 990 

ae Loe ΤΟ 2.5 152 
Pe 20. 19 335 

2 Ki. (4K.)1.6,16 152 
ie Zeal 3. 395 
See - 335 
ORT . 476 

ιν 21. 3 ey D2 

Aen 2 TOTS s 122 

1 Chr. 9. 25 . 449 
» [1.18 448, 449 

ro. ME Gy . 450 
24. A 439 

2 Chr. Te 450 
ye " 450 
22 10 476 

9 29.21,27,30 450 

cee eae aie 450 

» 34. 13 439 

=p, Syn 21 450 
Τῆι 55 att 454 
Esth. 1. 6 477 
lO 450 

veh Os, I 450 
Job 3. 21 251 


So with 

PAGE 

JODr4. 13) 126 
my 12. 5 (δὴ 222 
Sa αν τς 459 
SL Saal Aes 459 
mp 21. 2 - 102 
ee ΗΟ 459 
IP saat al 5 458 
=p BS : 379 
» 16 (15): 8 190 
» 25 (24). 14 449 
5 32. (3m) ar 190 
» 32 (31). 9 449 
» 48 (47). ὃ 189 
3) 49) (48)5 13. 2 289 
OB (02) 2 251 
9, 05 (64). 8 126 
» 76 (75). ὃ 461 
», 83 (82). 10 149 
» 94 (93) - 174 
3) 104. 210°. . 475 
» 118 (117).22 . 423 
LOS Ole 463 
.» 122 {121}. 5 458 
143 (142). 2 434 
Prov. 3.6 274 
ἜΣ τς 475 

Ay oe BSc 475 
Οὐδ: 86 

aA. Tillis Bee 274 

» 24. 5, 30 (45) 341 
6 1B au 196 
26. ὃ - 205 
Eccles. 6 ας 449 
Cant. 5. a 477 
Isa. I. 9 190 
ΠΣ, δ: 485 
ἈΝ A Set τς 228 
π᾿ pOoe. mie 424 
~» ChB 472 
Ὁ LO La 110 
ay) ZONZONe 442 
i PSE - 5 ar 
93 37-17-20) Ὁ SG 
A, ΠῚ, τ 158, ἅτ 
32. 3 . 189 
» 44. 16 ead 
AGS 206 
» 54. 13 459 
τ 29 408 
» 59. 13 246 
mp ep # 123 
Jer. 2. 27 124 
2. Τῇ 475 
Ae hi 152 
22, 22 469 


502 INDEX TO QUOTATIONS. 


PAGE PAGE PAGE 
Jer. 23. 16. 395 | Dan. 3.19 . 439, 450 | Hos. to. 14. . 423 
-ν Ὁ 2 5. zouge. 12) 400 τῷ ον: . 428 14. I . 423 
οὐ 321(89) a See 89 τ Στὸ - 450 Joel 4 (3). 18 . 475 
yo 136 (41. 395. τ Oe » 3 . 448 | Jon. > ἘΣ 16% 7.448 


4. 
5. 
ise 
» 30 (45).6 . 363 Ol. . 446 > : . 442 
Lam. 1. 16 . 4562 » ©6.14(15) . 446] Mic. ᾿ : (2) . 449 
Ezek. 9.9 . . 442 ap. (puistiec olson Nahas 10). 423 
τῆς . 440 ΕΣ ΤΕ 225 ον 565. ἜΓΑ .Ἐ2:.2.. - Zuo 
see AA . 440 » 6.23 (24) .. 446 Spe eee oat . 189 
Dan. 1.16. . 452 sss IO} 28 169) |" Zechi7. 3): . 447 
eee .. Γ4 50 Sn Veh . 441 op  CArites Ἐ 20 
Ἔα, ΤΣ ς . 446 mm e252 ἈΚ Fee ee . 419 
ay ee 22.. : - 475 ΣΕ ΤΩ: ἢ 435 ᾿ς L212 . 439 
mo. Be Sys . 430 | Hos. 9. 13 449 BA LTA Sh. . 475 
Pee Ore . 450 


PAGE PAGE PAGE 


1 Esd. 3. 5,9 5 CEG |) Sit, τῶ. Ae) - - 162 4 Ezra 4. 26 . 444 
Mee 4-30 5, PAWS || in, 23:12 πο} : “412 ae 4525 . 435 
st 4.2} τ 80) τ 32. τ {71.21) 425 λα 437 . 444 
ἌΓ 4. 52,63... 495}}) 26.033)» τ 5 aly - 5. 30 . 444 
ἘΝ 5-100 τ OOM μια 60. 20! - ~ BY, 5 Be 3 . 445 
6: 82 . 435 | Sus. 56 : eee OO ὅς 6. 14, 20..." 435 
a On 48 14} ΠΟ δι. 8... . 422 8.5 Ὁ . 445 
2 Esd. 7. 22 5 1.30 AO ANNs . 430 - 9. 41 . 445 
10. τῶ . 460 A Seto 5 495 τ: 10. 19 . 445 
Ep. Jer. 72 477} 3. πὲ τὸ. τὸ B 177 sam ΤΙ 37 454 
Jdth. 4. 2 442 ἜΣ Ἐπ 25 16, - 466 s 12. 40 454 
pens 11 0) 192 Ἂν MG ὦ - 362 13. 26 . 435 
1 Mae. Io. 8 . 445 4. 4 . 147| Test. XI. Patr— 
ΕΣ 10. 27 . 466 | Wis. 2.19 . . 405 Test. Reub. 4.1 211 
ΑΞ 10. 89 . 138 sn On ie : 99 my Coe) ic 67 
2 Mae. 7. 27 . 800 Ih AD e298 53 Os kee e203 
Sime 12. 27 oats || ἡ ἴθι Th 5 > ils Test. Sim. 4.5. 211 
4 Mac. 12. 7 SoD ὙΠ OS Test. Benj. 7.4 - Lid 
y 16. 15 Suara on £4 eee Ae 20. . 444 | Enoch 89. 42-49. 456 
Sir. 6. 7 : - 198 » 4.4. . 435 


(4) Harty CHRISTIAN WRITINGS. 


Clement of Rome (i/a.D.) 


PAGE PAGE PAGE 
ad Cor.12.7 . 445 | ad Cor. 20. 9 - 358 | ad Cor. 28. 3 . 344 

14.4 VeyeioGo 21.9 . 435 

Ignatius (ii/a.D.) 

Eph. 3.2 . 445 MAGn γ. τ. 4178 Philad.6.2;10.1 473 
31 - 473 ΠΟ ΤΡ 405 Polyc. ad init. . 106 
18. I 365 ΤΙ Ww 5 969 To Fo NGS 

20.01 346 ΤΥΟΙ͂Σ 2. 3} a BP 


Didache (ii/a.p.) 


2. ἢ- . 434 1Ιθ. 5 τ ὩΣ 
Baie . 366 τὸ. δ΄ te aloe 


INDEX TO QUOTATIONS. 503 


Barnabas (i/ii A.D.) 


PAGE PAGE PAGE 
4.10 406, 473 TZ Oe 
LOMET 5. es 205 2. 900 


Shepherd of Hermas (ii/a.D.) 


Vass τ τ 2417 Mand. ἵν. 2.1. 407 Sim. v. 1.1 358, 466 
Mb dis wt Gg 340 Wa Zo lint) 491 VWs Ue 5 5.2.0, 9.0 
We, 5 Lite ὙΠ 2, τ 95Ὲ Pel .- 910 
1| τοῦ 7, og UGS sae os 96 1X) 10515) 10 00) 
1|| 1.2: 275 ΧΙ 1. 2,4. 910 1X. 21:3 997 
eee 6 OEE Sim. iii. 1.4 . 346 ἸΧ 20 1. 910 


ix. 26.3 . 406 


Martyrdom of Polycarp (ii/a.D.) 
5.1; 16.1 346 | 18.3) 2 1s04. | ne . 398 


Epistle to Diognetus (ii/a.pD.) 
Be 415] 6.92. 372) | 12.9 . 407 


Justin Martyr (ii/a.p.) 
Apologyi.15,16 442 | Ape: 35. - 109.) Apo 8 37 . 362 


Gospel of Peter (ii/a.p.). (Roman numerals mark J. A. Robinson’s 
sections, Arabic figures, Harnack’s verses). 


ii. 3 : . 383 vi. 24 . 387 
lil. 7 ς - 409 x. 40 . 388 


Apocalypse of Peter (ii/A.p.). (Preuschen’s Antilegomena.) 
4. ὃ . 445 


Book of James (ii/a.p.). (Protevangelium Jacobi.) 
vi. I c . 434 


Pseudo-Nicodemus, Acts of Pilate (? iv/a.p.). (Tischendorf’s Ev. Apocr. 
1B) 


ero we 0 


Acts of Barnabas (iv/v 4.D.). (Tischendorf’s Act. Apost. Apocr. p. 60.) 
Bg : . 427 


(e) INScRIPTIONS AND OsTRACA. 


Archiv 
Archiv fiir Papyrusforschung, ed. U. Wilcken. 
PAGE PAGE 
1h 20) - - 170 | v. 163 ε - PAY 
Audollent 


Defixionum Tabellae, ed. A. Audollent (Paris, 1904). 
no. 92 Ξ . 213 


BCH 
Bulletin de Correspondence Hellénique. 
xvii. Ὁ: 78 . - 3807 


504 INDEX TO QUOTATIONS. 


Cagnat 
Inscriptiones Graecae ad Res Romanas pertinentes, ed. R. Cagnat (Paris 
1911- 
PAGE PAGE PAGE 
i. p. 26 (no. 41) - 79 | iii. p. 75 (no. 159) 437 | iii. Ὁ. 427 (no. 1145) 70 
GDI 


Sammlung der griechischen Dialektinschriften, ed. H. Collitz and F. Bechtel, 
(Gottingen, 1884— ). 


2561 D47 . ace 
IG 
Inscriptiones Graecae, ed. cons. et auct. Acad. Regiae Borussicae (Berlin, 
1815. 


IGSI (=IG vol. xiv.) 
Inscriptiones Graecae Siciliae et Italiae, ed. G. Kaibel (Berlin, 1890). 
no.966 . se HS |"no. 167% ~~ . 349 


IMAe (=IG vol. xii.) 


i. 406 : - 106 | iii. 248 ς - 127 
li. 168 ξ ΕΒ δ nik Gah. Bio) < - 124 | v. 653 > . “80 


JHS 
Journal of Hellenic Studies (Hellenic Society). 
iv. 385 : - 126 


Kaibel 


Epigrammata Graeca ex lapidibus conlecta, ed. G. Kaibel (Berlin, 1878). 
no. 426 - . 140 


Larfield 


Handbuch der Epigraphik, W. Larfeld (Leipzig, i. 1907, ii. 1902). 
i, ἜΘ᾽ - ξ . 


Letr. 


Recueil des inscriptions grecques et latines de 1 Egypte, ed. M. Letronne (Paris, 
i. 1842, ii. 1848). 


LIED Ὁ : oe Ὁ.91.}|. 025 : . 189 
Magn. 
Die Inschriften von Magnesia am Maeander, ed. O. Kern (Berlin, 1900). 
no. 17 : - 195 | no. 47 : 95 
OGiS 


Orientis Graeci Inscriptiones Selectae, ed. W. Dittenberger (Leipzig, 1903-5). 


no. 193 : - 66] no. 201 . 202, 380, 425 ) no. 214 : = 00 
194 - . SY) 


Perg. 


Die Inschriften von Pergamon (in Altertiimer von Pergamon, viii.), ed. 
M. Frankel (Berlin, 1900). 


no. 248 Ε <P 


INDEX TO QUOTATIONS. 505 


Preisigke, Sammelbuch 


Sammelbuch griechischer Urkunden aus Agypten, ed. F. Preisigke ΠΣ 
1915- ).Ψ 


PAGE PAGE PAGE 
Noy E3240 0: . 264 | no. 1540 . ᾿ | 154 


PAS 


Papers of American School of Classical Studies at Athens (Boston). 
lii. 204 177 


Ramsay, C. and B. 
Cities and Bishoprics of Phrygia, by W. M. Ramsay, 2 vols. (Oxford, 1895, 
1897). 
ii. p. 662 (no. 627) 272 | ii. p. 565 (no. 466) 460 


REGr. 

Revue des Etudes grecques (Paris, 1888 ff.). 
Xvlil. 205. eae 
Syll. 


Sylloge Inscriptionum Graecarum,. ed. W. Dittenberger (Leipzig, ed. 2, 
1898-1901, ed. 3, 1915-1924). 


ed.? 177 (=ed.° 344) 66, | ed.2 (=ed.® ) ed.?737 (ed.21109) 101, 
200 467 
493 ( 722) «Ἰὴ8 596 ( IOII) 178 757 (i vetr25)) αὐδὲ 
583 ( 996) 7298, 598 ( 1012) 178 803 ( 1169) 89 
907 
588 ( om.) 328 615 ( 1024) 299 805 ( itr), 1592 
594 ( 1009) 178 641 ( 1044) 386 gol ( 1251) 108 


Viereck SG 


Sermo Graecus quo Senatus Populusque Romanus . . . usi sunt, by P. Viereck 
(Gottingen, 1888). 
p. 16. : . 199 


Meyer Ostr. 


Ostraka der Sammlung Deissmann, ed. P. M. Meyer in Griechische Texte aus 
Agypten (Berlin, 1916). 


no. 57 : . 380 


Preisigke Ostr. 


Die Prinz-Joachim-Ostraka, edd. F. Preisigke and W. Spielenberg (Strassburg, 
1914). 


no. 15 . 114 


Wilcken Ostr. 
Griechische Ostraka, ed. U. Wilcken (2 vols. Leipzig, 1899). 


LP LOOM. ‘ 2 900] 1 292 : . 339 | ii. 1089 ‘ 5 73 
ΤΣ. - : 999 | ii. 1084 73 


Mélanges Nicole 


Recueil de Mémoires de Philologie, Classique et Φ Archéologie offerts ἃ Jules 
Nicole (Geneva, 1905). 


p. 184 : τ 102 Ὁ. 185 : . 210 


506 INDEX TO QUOTATIONS. 


(f) Papyrt. 


N.B.—For the references given in brackets, see pp. 511f. of this Index. 
Wilcken and Mitteis’s Chrestomathie is cited by volume and page. 


P Amh 
The Amherst Papyri,i. ii., ed. B. P. Grenfell and A. 8. Hunt (London, 1900-1). 


PAGE _ PAGE PAGE 
ii. 37 (Witk. p. 93) 189 | ii. 75 5 oe LES | li. 93 (Chr. i. 373) . 189 
ii. 68 (Chr. i. 439). 189 


Archiv (see under (e) above) 


li. 126 ᾿ . 998.} ν.. 232 Ε oe 79 | vi. 104 . 344 
v. 38 . 286 


BGU 


Agyptische Urkunden aus den kéniglichen (staatlichen) Museen zu Berlin : 
Griechische Urkunden i.—vii. (Berlin, 1895-1926) Vol. i. nos. 1-361 (1895). 


no. 16(Chr.i. 145) 133 | no. 86 (Chr. ii. 350) 205 | no. 250(Chr.i. 114) 221 
67 282 


27(Chr. i. 525, 93 : : ον 101 
Sel. p. 101). 62 133 : 66, 170 326 (Chr. ii. 365) 206 
38 - - 42) 163 : ΠΟ 350 ὃ e205 


γι i FES 10 


Vol. ii. nos. 362-696 (1898). 


no. 385(Chr. i. 133) 91, | no. 427 5 . 443 | no. 538 A 5 ivy 

127, 425 511 (Chr.i. 26) 119 602 Ὁ a PAV 

388(Chr. ii. 107) 261 523 : . 432 647 - - 89 
422 ΠΕ 177 


Vol. iii. nos. 697-1012 (1903). 
no. 781 : . 347 | no. 883 44, 113, 144 | no. 916 ς - 66 


785 - 109 896 : - OH 948 : aA 
816 δ 6. Hy 907 : - 349 975 5 οἰ 99 
830 ὃ ΕΠ ΘΙΟ - 5 SDI 
846 (Sel. p. 94) 136, gI2 . 5 
429 
Vol. iv. nos. 1013-1209 (1912). 

ΠΟΙ ΘΙ “ἡ: OOS Sno ΕΤΘΙ Ὁ Ὁ . 432 ΠΟ. 1134 . BPA 
τσ Ne 5 eH | Τα . 205, 374 ΤΙ: . 432 
1079 (Chr. i. 85, 1117 (Chr. ii. 129) ΤΟ  : ΠΡ) 

Sel. p. 40) 31 111, 310 Το τς - SOG 
LO) as . 432 1126 . 9305 LO Omens . %r14 
MIO2 0 ats . 347 ΕΓ 7a . 204 PLOAR ὅν etal 
LOS ἢ zee 1130 0S 2900; 355 
P Catt 


P. Cattoui, ed. G. Botti in Revista Egiziana, vi. 529 ff. See Chr. ii. 420 ff. 
col. i. Σ : one 69 


CP Herm 
Corpus Papyrorum Hermopolitanorum, i., ed. C. Wessely (Leipzig, 1905). 
no.6. : . 202 | no. 101 : - 374 | no. 125 3 . 305 
73 β eo: 102 (Chr. i. 346) 86 127 «Ὁ 109. 955 
83 Ξ . 3714 119 : . 374 


94 ᾿ . 814 121 . - 305d 


INDEX TO QUOTATIONS. 507 


CPR 
Corpus Papyrorum Raineri, ed. C. Wessely (Vienna, 1895). 
PAGE PAGE PAGE 
no. I (Chr. ii. 240) 90 | no. 175 j : 2d | σἷοὲ 252 5 66, 170 
21 Ξ < (07 | 220 A a) 266 ὰ : 
P Fay 
Faytim Towns and their Papyri (1900), 
no. 82 é . 174 | no. 108 . . 415 | no. 347 ὃ ὌΠ 9 
83 : τι τ: 109 450 | 
P Flor 


Papiri Fiorentini, i.—iii., edd. G. Vitelli and D. Comparetti (Milan, 1906-15). 
Vol. i. nos. 1-105 (1906). 


no. I (Chr. ii. 274) 66 | no. 61 (Chr. ii. 89) 66 127 : ate 
37 : - 86 (Chr. ii. 282) 66 
Vol. ii. nos. 106-278 (1908). 
no. 160 = . 347 | no. 175 5 ON ΠῸΣ ΤᾺ ς οὐ Ὁ 19 
170 : . 869 176 < . 256 262 . . 14] 
P Gen 


Les Papyrus de Genéve, i., ed. J. Nicole (Geneva, 1896-1900). 
no. 24 3 a) 0] πὸ 54 : - 108 | no. 74 A = 101 


P Giss 


Griechische Papyri zu Giessen, i., edd. O. Eger, E. Kornemann, and P. M. 
Meyer (Leipzig, 1910-1912). 


no. 10 : a LO 1 πὸ: 49 ἃ : 199 | no. 79 ὦ : 70 
12 ὃ 5 τί 54 Ξ 5 Pals 86 δ «70 
40 - . 224 

P Grenf i. 


An Alexandrian Erotic Fragment, and other Greek Papyri, chiefly Ptolemaic, 
ed. B. P. Grenfell (Oxford, 1896). 


HOA Ty : . 275 | πο. 53 (Chr. 1. 158) 445 


P Grenf ii. 


New Classical Fragments, and other Greek and Latin Paypri, edd. Β, P. Grenfell 
and A. 5. Hunt (Oxford, 1897). 


no. 36 (Witk. p. 120) 221 | no. 74 : a 147 
P Gurob 


Greek Papyri from Gurob (“ Cunningham Memoirs,” no. xii.; see under P Petr), 
ed. J. G. Smyly (Dublin, 1921). 
no. 32 . « 0342 


P Hamb 


Griechische Papyrusurkunden zu Hamburg, i., ed. P. M. Meyer (Leipzig, 
Berlin, 1924). 
no. 4. 3 = + 42) | nos τὸ A 70,347 | no. 89 Σ - 346 


P Hib 
The Hibeh Papyri, i., edd. B. P. Grenfell and A. 8. Hunt (London, 1906). 
no. 53 - . - 467 
VOL. II. PART III.—34 


508 INDEX TO QUOTATIONS. 


P Iand 


Papyri Iandanae, parts 1-4, edd. E. Schafer, L. Eisner, L. Spohr, and G. 
Spiess (Leipzig, 1912-1914). 

PAGE PAGE PAGE 

no. ὃ. . « 433 | no.9. . 22431 |) τὸ: 62 - - 05 


P Leid 


Papyri Graeci Musei antiquariae publici Lugduni-Batavi, 2 vols., ed. C. 
Leemans (1843, 1885). 


Gy » 161162 $6 (UPZ106) ©". 89: We co eee 


P Lille 


Papyrus grecs de Lille, i., edd. P. Jouguet, P. Collart, and J. Lesquier (Paris, 
1907-28). 


no. 5 . « 83,85 | no. 11 : - 221 | no. 17 (Witk. p. 28) 83 


P Lips 


Griechische Urkunden der Papyrussammlung zu Leipzig, i., ed. L. Mitteis 
(Leipzig, 1906). 


no. 69 " . 346 | no. 87 : . 174 | no. 121(Chr. i. 205) 72 


P Lond 


Greek Papyri in athe British Museum, vols. i. and ii., ed. F. G. Kenyon; 
vol. iii., edd. G. Kenyon and H. I. Bell; vols. iv. and v., ed. H. I. 
Bell (London, τοῖς 1917). 


Vol. i. nos. 1-138. 


no. 22 (p. 7)(UPZ 17) 60 21) (0: ττῶὴ - 979 πὸ: 190 (ps4) Σ᾿ 209 
23 (p. 38) (UPZ 14) 122 (0: 122) 20 131 (pp. 166-188) 
110 130, 162 


Vol. ii. nos. 139-484. 


no. I9I (p. 265) 102, 103 | no. ae 177) a 172 
7. 11 TiO: 


Vol. iii. nos. 485-1331. 
no. 924 (p. 134) 
(Chr. i. 418). 150 


1159 (p. 113) 
(Chris 493) = 272 


no. 1170 (p. 196) 133 | no. 1178 (p. 216) 
II7I (p. 178) 270 (Chr. i. 185, 
1177 (p. 183) Sel. p.99) - 127 
(Chr.1.277). 344 


P Oxy 


Oxyrhynchus Papyri, vols. i—vi., edd. B. P. Grenfell and A. 5. Hunt; vols. 
vil.ix., ed. A. 5. Hunt; vols. x.—xvi.,e dd. B. P. Grenfell, A. S. "Hunt, 
and H. I. Bell (London, 1898-1924). 


Vol. i. nos. 1-207. 


ΠΟΣῚ «ὁ 80, 191, 409 | no. 63 272 | no. 117 - 22p2450 


Gee - . 444 92 5 5 212 110 (Sel. p. 102) 215 
33 (Chr. i. 35) IOI - 5 00 
109, 206 113 = 992 


38 (Chr. ii. 65, 
Sel. p. 53) . 189 


INDEX TO QUOTATIONS. 509 


Vol. ii. nos. 208-400. 


PAGE PAGE PAGE 
no. 208 ‘ ; 81) no. 248 : » 1} [ὦ 29 (ΟἿΣ Ἢ, 
209 4 75, 81 252 Ξ . 189 262). ay 60 
237 - 110,280 265 : «“ 90 281 (ΟΝ 11. 
242 : 1095, ΤΉ] 188 


74) 
294 :. 09. 86, 151 
Vol. iii. nos. 401-653. 
no. 404 5 - 81] no. 502 132 | no. 513 (Chr. i. 217) 291 
| 531 (Chr. 1. 567) 70 
532 5 2 205 


406 ns 81 505 (Chr. ii. 307) 113 
478 (Chr. i. 253) 349 


Vol. iv. nos. 654-839. 
no. 654 = fi) | no. 657 . 1ῦ, 80, 83 


no. 744 (Sel. p. 32, 
655 4 72, 81 736 5 70, 127 Witk. p. 131) 63 
656 : 72, 81 
Vol. v. nos. 840-844. 
no. 840 5 81, 343 
Vol. vi. nos. 845-1006. 
no. 847 A ~ 8 
849 5 OL 
850 ὃ 81, 90 
889 : . 174 


Vol. vii. nos. 1007-1072. 


«-..ὕ......ὕ.. 


no. 896 (Chr. i. 73) 173 | no. 940 . 270, 440 
gi2 - 199 967 2 oO 
933 2 70, 87 986 ὃ . 205 
936 ᾿ 5 70 


ΠΟ. 1007/5 5 8] [πὸ Boyne) - 5 81] 1 Π0: ΙΘ΄ - . 432 
1008. 5 bill 1029) - - 161 TOON - . 203 
1009. 5 9. 1036: Σ > IS 1OO7 a - . 215 


Vol. viii. nos. 1073-1165. 


HO Ἰο.5 - . 8180. 1088 no. 1148. τς: 5 1 
1O7 Sus 5 lll 79, 106, 113, 132 Tali Ties : 8ῦ 
1079. - 51 1119 (Chr. i, ear: Πρ ἢ . 204 
1080. - OL 12 τ΄: 66 72 as . 80 


ivi) -: - 349 
Vol. ix. nos. 1166-1223. 


no. 1166 . ae SU NOW 1108" 5 till) toe, RK: og . 219 
LOZ 81 IAL ΝΣ OL 1204 . - 467 
Vol. x. nos. 1224-1350. 
No. 122450. τ 81] [πὸ 220 - ᾿ Ol Os Ε22Ο 2 - 80 
1225s 5 telll 122350 5 — Bil WAXY Ὁ 5 teal) 


Vol. xiv. nos. 1626-1777. 
no. 1676 « . 446 


Ρ Ῥας 
Paris Papyri, in Notices et Extraits, xviii. part 2, ed. Brunet de Presle (1865). 

no. 16 306 | no. 48 (UPZ 72) . 221) πο. 63 (ὉὀΡΖ 110) 224 

32 (UPZ 61, 49 (UPZ 62, 

Witk. p. 67) 222 Witk. p. 71). 427 

40 (UPZ 12). 110 51 (UPZ 78, 

46 (UPZ 71, Sel. p. 19) 70, 443 

Witk. p. 86) 222 
P Par 574 


The Paris Magical Papyrus, ed. C. Wessely, in Denkschriften der philosophisch- 
historischen Classe der kaiserlichen Academie der Wissenschaften zu Wien, 
xxXvi. (1888), pp. 75 ff. (=Wessely, Zauberpap.). 

line 86 : . 406 | line 2838 , 89, 202 | line 3037 . 286 
274. ς . 408 3007 - . 406 


510 INDEX TO QUOTATIONS. 


P Petr 
Flinders Petrie Papyri, edd. J. P. Mahaffy and J. G. Smyly (in Proc. Royal 
Trish Acad., ‘Cunningham Memoirs,” nos. viii., ix., xi. ; 3 parts, Dublin, 
1891-1905). 
Parti. (J. P. M. 1891). 
PAGE PAGE PAGE 


no. 29 (p. 78) (Witk. 
ὍΣ 32): ΚΣ 
Part ii. (J. P. M. 1893-4). 
no. 4 (7) (p. 11) . 221 | no. 13 (19) (p. 45) no. 23 (3) (p.70). 60 
4 (9) (p. 12) . 221 (Witk. p. 19) 428 35 (p. 115). . 349 
4 (11) (p. 14). 147 13 (19) (p. 45) 45 (ili) (p. 147) 107 


(Witk. p. 20) 459 
Part iii. (J. P. M. and J. G. 5. 1905). 
no. 53 (πη) (p. 153) no. 53 (p) (Ρ. 154) 91 
(Witk. p. 44) 63 | 
Preisigke, Sammelbuch (see under (e) above). 
no. 2632 - > “162' [ mo. 4317 Ὁ: τ 413" | mo. 5224 2 Ree ji 


P Rein 
Papyrus grecs et démotiques, ed. Th. Reinach (Paris, 1905). 
no. II Ξ τ {δ 


P Ryl 
Catalogue of the Greek Papyri in the John Rylands Library, Manchester, 
2 vols., edd. A. S. Hunt, J. de M. Johnson, and V. Martin (Manchester, 
1911-15). 
Vol..i. (A: 5. H. 1911). 
no. 29 ὁ a 79 
Vol. ii. (J. de M. J., V. Μ., and Δ. 5. H. 1915). 
no. 90 Β - 15 7 πὸ ΤΡΖ. Ξ :- ΟΠ πος 221 - .: 9] 


113 3 See 157 ς - ΠῚ 233 5 10 
127 Ἔ 5g) 160 A 5 1s 239 - 346 
130 3 oy OL 202 4 - 169 441 ὃ ane stile 
153 - 106, 429 209 Σ = O09 

PSI 


Papiri Greci e Latini della Societa Italiana, i—viii. (Florence, 1912-27). 
13h Ὁ . 346 | iv. 339 - : 51] 1 ἢν; 435 A . 409 
ili. 175 - soto ἵν. 359 Ξ - 351 | vi. 600 : - 346 
iv. 338 = OL: 


P Strass 
Griechische Papyrus zu Strassburg, ed. F. Preisigke (Leipzig, i. 1912, ii. 
1920). 
15 2a - oll 
P Tebt 
The Tebtunis Papyri (University of California Publications), 2 vols. (London, 
1902-7). 


Vol. i, edd. B. P. Grenfell, A. 5. Hunt, and J. G. Smyly. (nos. 1-264.) 
no. 20 § . 395 | no. 104 (Chr. ii. 320) 113 | no. 114 - = 90 
33 (Chr. i. 9, 110 ξ 166 119 A ~> 98 
Sel. p. 30) . 88 . 
61 113, 220, 224 


INDEX TO QUOTATIONS. δ11 


Vol. ii., edd. B. P. Grenfell, A. S. Hunt, and E. J. Goodspeed. (nos. 265-689,) 


PAGE PAGE PAGE 

no. 265 ἐ 00) πο 2110 7. 1 134) 204 | no. 421 (Sel. p. 107) [ΕΣ 
272 : - 125 388 66 472 : 163 
295 Ξ = 105 409 5 = 910 Ostr. 6. :- 66 


P Thead 
Papyrus de Théadelphie, ed. P. Jouguet (Paris, 1911). 
no. ti. 5 . 349 


P Tor 
Papyri Graect Regit Taurinensis Musei Aegyptii, 2 vols., ed. A. Peyron 
(Turin, 1826-7). 
no. 13 (UPZ 118) 
(Chr. ii. 26) . 139 


UPZ 
Urkunden der Piolemderzeit (altere Funde), i., ed. U. Wilcken (Berlin and 
Leipzig, 1922-7). 
no. 12 PPar4o . 110 | no.62 PPar49 . 427 | no. 106-P Leid G.: 89 


14 P Lond i. 23 110 ΠΕ ee ANG). PP 110 P Par63 224 
17 P Londi. 22 60 72 P Paras. 991 1¥8 P Tor13~ 139 
Or PsPar 22 ς 2.9.9 78 P Parsi . 443 


Chrest. i. and ii. (also, in this index, Chr.) 
Grundziige und Chrestomathie der Papyruskunde, i. 2, ed. U. Wilcken, 
ii. 2, ed. L. Mitteis (Leipzig and Berlin, 1912). 


Vol. i. 
no. 3PTebti.33 88 | no. 156 P Lond iii. no. 314 P Amhii. 93 189 
14 BGU ii. 511 119 1178 6 IBY 355 - Lond iii. 
48 P Oxy ii. 896173 173 PLipsi.121 72 924 . 150 
60 BGU iv. TOS cb O)Xayau ite 374 P Amh ii. 68 189 
1079 5 9] 513 . 291 397 P Oxy viii. 
87 BGUi.250 221 193 P Lond iii. 1110 . 221 
100 BGU ii. 385 1177 . 344 415 P Lond ii. 
91, 127, 425 218  Ῥ Οχυ 1. 1159 sae 
114 BGUi. 16 . 133 478 . 349 445 BGUi.27 . 62 
131 P Grenf. i. 296C P Herm 482 P Oxy ii. 
SS ine olds 102 5 tht} 531 oY 
Vol. ii. 
no; ΖΘ ΟΡ. 139) |mos 15 Ῥ- ΠΗ, 1] no. 247 P Flori. 61 66 
58 P Oxy i. 38 189 333 . 204 285 P Tebt 1. 
66 P Oxy ii. 281 188 181 P Lond ii. | 104 = Wits} 
80 P Flori. 61 66 262 66, 172 306 BGU i. 86 . 205 
οἱ BGU ii. 388 261 220 CPR 1 = 90 316 BGU i. 326 206 
107 BGU iv. 1117 25ΘΝ ΟΣ 11: 350 P Oxy iii 
111, 310 270 ~ te) 505 . 113 


243) P Hlorist .. 66 372 P Catt col. i. 69 


Selections (also, in this index, Sel.). 
Selections from the Greek Papyri, ed. G. Milligan (Cambridge, 1910). 
no. OPPars1 . 443 |no. 19 P Oxyi. 38 189|no. 41 BGUi.27. 


7) 
ok 


11 P Tebti. 33 88 37 BGU iii. 846 42 P Oxyi. 119 215 
12} Oxy iv. 136, 429 44P Tebt ii 

744 = (a; 40 P Lond iii. 421 . 444 
15 BGU iv 1178 - 127 . - 


1079 ». dl 


512 INDEX TO QUOTATIONS. 


Witk. 


Epistulae Privatae Graecae quae in papyris aetatis Lagidarum servantur, 
ed. 8. Witkowski (2nd edition, Leipzig, 1911). 


PAGE PAGE PAGE 
πο: 8.2) Retr ou: no. 27P Petr iii. no. 64P Grenf ii. 
13 (19) 428, 459 5S (n) 63 30. « : 591 
16P Lillei. 17 83 38 PPar4g . 427 72. Oxy iv. 
18 P Petri. 29 270 50 P Amhii. 37 189 744 - 63 
(9) Greek LITERATURE. 
i. Classical. 
Homer (? viii/B.c.) 
PAGE PAGE PAGE 
Iliad ii. 545 - 66 | Odysseyii. 114 . 429 | Odyssey xxiii. 110 460 
iv. 422 - 501 vi. 59 - 0. 
Vil. 475 . 286 xii. 73 . 425 
Hesiod (Ὁ x/Vviii B.c.) 
Scut. 348 . 71 
Mimnermus (vii/B.c.). (Bergk*, Poetae Lyrici Graeci.) 
JO Tate - . 106 
(2?) Epimenides (vi/B.c.). (Diels, Fragm. der Vorsokratiker, ii. 185 ff.) 
9, 137, 158 
Pindar (v/s.c.). (Bergk‘.) 
PPE stays ee LOZ 
Aeschylus (v/B.c.). 
Supp. 248 - 386 
Sophocles (v/s.c.). (Fragments cited from Pearson.) 
Aja-*54. . 451 | Electra 343 . . 459 | Philoct. 123 . - 910 
807 : - 459 1207. : . 421 gu <x - - 435 
Ant 51 . . 272 | Oed. Tyr. 650 - 421 | Frag.201 .- . 440 
TOT ὁπ δ ς . 468 
Euripides (ν [Β.0.). (Fragments cited from Nauck, ed. 2.) 
Medea 1398 . 423 | Orestes497 . -. 459 | Telephusfr.714 . 442 


Aristophanes (v/B.c.) 
Acharn. 687 . 361 | Nubes4io . . 246 | Vespae 213 . . 442 
Lysistr. 1000. 461 | Ranae 35,37 . 210 TOOT ou ule - 498 


Hippocrates (v/B.c.) 
Art. 824 . - 389 


Herodotus (v/B.c.) 


L326 “Se ee U2 i inh agG etn : 4409) viniege: SP gatos 
11: 22 - oe 80} ν 8 - ΠΡ 98} 0} τοῦ oaths} 


INDEX TO QUOTATIONS. 513 


Thucydides (v/s.c.) 
PAGE PAGE PAGE 
151 . - 450. Ὁ. Ὁ = - 22) | aves 100 ~ . 428 
Sy Sa 8 05) 1 mes 5 - -- 159 


Batrachomyomachia (v/B.c.) 
θη. « ~ ᾿Ξ] 


Xenophon (iv/s.c.) 
Anabasis— Cyrop. τι. iii. 8 . 428 | Cyrop. Υ. iii. 30 . 428 
I. viii. 12 . 459 TO) es . 460 ΜΝ. 2 Ours . 428 
WG IKe - 422 Ti πη νι, ee 128 Vb it Sis τ . 466 
Ve vi: 12 . 433 Ts 1: 38)  - . 435 VALE 1 16 τ . 459 


ὙΠ. vis 32 . 409 ἘΠῚ sie) 5 . 428 VIII. ii. 24 . 428 
Cyropaedia— INAS ws | . 435 | Hellen. r.ii.10 . 468 
I tb ς . 428 τν TL, WO) o . 428 Ν 11:20. - . 433 
Teves wae a alt ἜΝ: 11. 172 . 408 ὙΠ: ae ΤΟ - . 460 
Ἐν Τὸ re a Es Naot 1G) ς . 435 | Memorabiliam.i.1 421 


Plato Comicus (v/iv B.c.) 


Zi 5 5 900 
Plato (iv/B.c.) 
Cratylus 4046 . 129 | Hippias Major— Phaedrus 274d . 145 


Euthyphro2a . 436 301be - . 303 | Republici. 331¢ . 428 
Gorgias 521d . 421 | Lawsviii.839a . 303 | Theaetetus144c . 428 
Phaedo 59e . . 422 
Andocides (v/iv B.c.) 
de Myst.89 . 468 


Lysias (v/iv 8.0.) 
πεῖν, ἡ ς- . 459 


Isocrates (iv/B.C.) 
iv. 4 : - 459 
Demosthenes (iv/B.c.). (Cited by orations and Bekker’s sections. Reiske’s 
pages in brackets.) 


iii Aphob.1(844) 459 | Lacr. 32 (933) . 347 | iOnet. 33 (873) . 433 
Fals. Leg. 278 Meid. 93 (544) . 440 Timocr.7 (702) . 194 
(430) . - 407 


Lycurgus (iv/B.c.) 
Ws) 6 : . 288 


Aristotle (iv /8.c.) 
Metaph. ix.4 . 166 | Polit. viii.3.10 . 154 


il. Hellenistic. 


Aratus (iv/iii B.c.). (Ed. Bekker.) 
PAGE 
Phaenomena 5 . 9 


Cleanthes (iv/iii B.c.). (Cited from von Arnim, Stoicorum Veterum Frag- 
menta. ) 


Ie Gye ὁ τὸ 9 


514 


Menander (iv/iii B.c.). 


ἘΠῚ 75 


Theocritus (iii/B.c.) 
iv. 35 


Callimachus (iii/B.c.). 


INDEX TO QUOTATIONS. 


PAGE 
9 


69 


Epigr. 43(s), 42(M) 435 


Polybius (ii/B.c.). 
153 
li. 41. 15 
Mi Age 
Wes 7) 


Dionysius Halicarnassensis (i/8.c.). 


Ant Ἦ: 2 ΤΖ 


Meleager (i/B.c.). 
δ5 Ὁ 


Strabo (i B.c./i A.D 
v. p. 230 


Philo Judaeus (i/A.D.). 


De Agric. 19 (i. 

p- 314M) : 
De Mutat. Nom. 
Q (i. p. 588M) . 


Flavius Josephus (i/.p.). 


Antig.— 
πὴ 6: δ (37) 
vi. 13. 7 (305) 
vill. 3. 3 (72) . 


Plutarch (i/A..). 


Ages. ΧΙ]. 5 
Cicexx 2 


Epictetus (i/ii a.D.). 


Enchiridion 12 . 


Arrian (ii/ A.D.) 
Anabasis i. 3. 2 
Epicteti Disser- 

tationes— 
rh Oh tls 
16 2 Sie 


Aelian (ii/4.D.) 
De Nat. Anim. 
i. 19, 48,55 . 


δ 


(Fragmenta, ed. Meineke.) 
PAGE 


(S=O. Schneider’s ed., M=Meineke’s ed.) 


(Cited by book, chapter and section.) 


459 | ii. 47. 1 

355 | ii. 50. 10 
300) |) ἢ. 5455 

354 


343 | Ant. BR. 4. 24 


128 


432 | vil. p. 319 


De Somn. 13 (i. p. 


365 632M) : 
126 


Antig.— 


152 xvi. 6. 4 (168) . 
445 Xviil. 5. 2 (118). 
152 XVili. 9. 7 (363). 


392 | ili. 44.°5 
343 | iv. 19. 10 
340 | xxxvili. 10. 10 


(R for Reiske’s pages.) 
361 | Lp. ad Pomp. 756R 


(Cited by Casaubon’s pages.) 


272 | xiv. p. 648 


(Mangey’s volume and page in brackets.) 
De Spec. Leg. (ii. 


278 p- 311M) 


(Niese’s sections in brackets.) 


Bell. Jud.— 
128 i. 20. I (390) 
445 v. 4. 2 (145) 
445 v. 12. 2 (505) 


PAGE 


380 
307 
400 


434 


85 


461 


445 
152 
152 


(The Lives cited by chapter and section, the Moralia by 
Xylander’s pages [Venice, ed. 1570].) 


346 , C. Grac. xvii. 4 
377 | Mor. 3368 


347 | 42 


202 | Epicteti Disser- 
tationes— 
i. 29. 49 
420 li. 16. 29 
436 li. 18. 24 


De Nat. Anim. 


425 | ii st. 


(See under Arrian.) 


400 | T. Grac. xxi. 2 
201 


ADS alesis 


Epicteti Disser- 
tationes—- 
420 ili. 22. 36 
380 1112. 10ς 
420 iv. 9. 13 


400 


436 


433 
433 
468 


INDEX TO QUOTATIONS. 515 


Appian (ii/a.p.) 
PAGE PAGE PAGE 
Hist. Rom. Praef. 
10 sere, 166 
Lucian (ii/A.p.) 
Bis Accusatus De Syria Dea Dialogi Mortuorum 
17 - 5 8k) |b 85 c : - PADIL } xa & : .. 425 
Marcus Aurelius (ii/A.p.) 
Vii. 3 > « 5840 
Vettius Valens (ii/A.p.). (Cited by page and line in W. Kroll’s ed., 1908.) 
το Πρ τη ον. Ἰυ}) 24Ou Tbe a ce τοῦ 


Artemidorus (11.Δ.}.} 


Oneirocritica— | 
1 12. πὰ . 901 [11 32:66 ‘ 5 yz) |) 1: Gat 5. ait) 


Athenaeus (ii/iii A.p.). (Cited by Casaubon’s pages.) 
AZAE ae ee 2S | OS7De Gs | Lames oOo 


Vitae Sophistarum 
li. 13 : 
Sextus Empiricus (iii/a.p.) 
li. 54 . . 166 | ix. 406 - GG 
Diogenes Laertius (iii/a.p.) 
W527 ee 129} υἱ. 2: 37 5 4 ots 
Marcellinus (iv/a.p.) 


Vita Thucydidis 
57 -  . 440 


Pseudo-Callisthenes (iv/a.p.). (Cited, after Helbing, from von Meusel.) 
li. 41 5 . 445 


Pseudo-Chion 


Epistolographi 
To: 30 τ΄ ΠΤ) 
Pseudo-Hippocrates 


Epistolae1o.6. 440 


Pelagia. (Legende der hl. Pelagia, ed. H. Usener.) 
ΧΕ 2 1 . 460 


Anthologia Palatina 
Vinlose. -  L66 | vate 72 τ TOT ston) ero eT O6 


Aesopi Fabulae. (F=de Furia’s collection, 1810; T=Teubner, ed. C. Halm, 
1889.) 
257(F), 284(T) . 477 
Scholia on— 


Homer, Od. xii. Demosth., c. Androt. 
336 : . 407 nes Ε . 444 


516 INDEX TO QUOTATIONS. 


(h) Latin. 


Varro. (Cited by book, chapter and section, Spengel’s numeration.) 
De Lingua Latina 
PAGE PAGE 
Ἰχ 2 γα κα 900 
Cicero 
Ad Att. xiii. 29. 2 390 


Catullus 

lzxxiv. tf =. ΘΟ, [-xeyat. <3 = 5. ee 106 
Virgil 

Aen. ix. 20. = 902 


Horace 
Carm.iv. 10.1 . 314 


Propertius 
He 25 12" - 434 


Persius 

Wie) 5O) 6 5 5 ΡΝ 
Martial 

Ibe US τις > 515 


Tacitus 
Ann. ii. 55. . 82 


Suetonius 
Calig. xxvii. . 47 


i) 


Il. INDEX OF WORDS AND FORMS. 


ΞΘ = 


(a) GREEK. 


a: origin 39—sound of 43—vowel 
changes 65 ff., 253 

-a as suffix 335 

ἀ- (=v privative) cpds. 280, 286— 
(=sm-) 285 

ἀβρασαξ (number of the year) 169 

ἄβυσσος : gender 122 

ἀγαθοεργῷ T5— -εργεῖν 91 

ἀγαθοποιία 83 

ἀγανακτέω 281 

ἀγάπη 335— -dw 384 

ayyapevw 67, 398 

ἀγγεῖον 344 

ἅγια, ἁγία 60 

ἄγκιστρον 369 

ἄγναφος 371 

ἁγνευτήριον 343 

ἀγνοεῖτε or ἀγνοεῖται 70 

ἄγνοια, -οέω 281, 390 

ἄγνωστος 276, 280, 370 

ἀγόραιος and ἀγοραῖος 57, 296, 337 

ἀγοράστρια 349 

ἀγορονομεῖον 344 

ἀγραυλέω 283 

ἀγρίδιον 346 

ἀγριέλαιος 276 

᾿Αγρίππας 119 

ἀγρυπνέω, -via 290 

-aywyéw: epds. in 388 

ἀγών 58, 136 

ἄδελφε δ8---ἀδελῴφός 285— -ότης 367 

ἀδημονέω 281, 391 

ans 22, 119 

᾿Αδραμυτηνός 106, 157 

ἀδύνατος 280, 370 

ἀεί and αἰεί 81, 163 

ἀετός (not αἰετός) 81 

-dfw: vbs. in 68, 404 

"Αζωτος 106 ἢ. 

ἀηδία 348 

ἀήρ 137, 363 

ἀθέμιτος 371 

ἀθερίξζω 326 

ἄθεσμος 353 

ἀθετέω 393 

ἀθροίξζω 407 

ἀθυμέω 308 

ἀθῷος 84 


αι: fusion with ε 80, and a 81 

αἰγιαλός 361 

ἀΐδιος 157, 336 

aidws 140 

αἰέν 136 

αἷμα 354 

αἱμαροοῦσα for aluo- 73 

αἱματοεκχυσία 271 

-αινα as suffix 348 

Aivéas 119 

αἴνεσις 374 

aivéw 386 

αἴνιγμα 354, 404 

-aivw: vbs. in 214, 401 

-aios, -atds, -acos as suffixes 337 

alperifw 408 

αἱρετικός 379 

-αίρω : vbs. in 214, 402 

αἰσχροκερδής 284 

αἰσχρολογία 284 

αἰσχύνη, -ύνω 358, 402 

αἴτιον 341 

αἰτίωμα 354 

αἰφνίδιος, ἐφν- 70, 157 

αἰχμάλωτος 271, 352, 370— -evw 400— 
-ἰζω 407 

αἰών 135 

αἰώνιος 157 

ἀκαιρέομαι 390 

ἄκανθα 118 

ἀκατάπαστος 87, 253, 317, 370 

᾿Ακελδαμάχ (-dx) 108, 109, 153 

ἀκέραιος 285 

ἀκμήν 163 

ἀκοή 335 

ἀκολουθέω 285 

ἀκροβυστία 277 

ἀκρογωνιαῖος 277 

ἀκροητήριον 342 

ἀκροθίνια 277, 341 

᾿Ακύλας 119 

ἀλάβαστρος 122 

ἅλας, dds 132, 136, 333 

ἁλεεῖς 76, 90, 142 

ἀλεκτροφωνία 271 

ἀλέκτωρ 365, 387 

᾿Αλέξανδρος 291, -δρινός (or -ivos) 359, 
-dpevs 350 


517 


518 


ἀληθής 289, -ινός 359, -evw 399 

ἀλήθω 383 

ἁλίξζω 408 

-ahis as suffix 362 

ἀλίσγημα 354, 387 

ἀλλά for εἰ μή 468 

ἀλλάσσω 403 

ἄλλῃ 84 

ἁλληλουιά 154 

ἄλλος 336 

ἀλλοτριοεπίσκοπος 63, 92, 272 

ἀλλότριος 369 

ἀλοᾶν 100 

-αλος as suffix 361 

ἁλυκός, ἁλικός 80, 377 

ἅλυσις 100, 374 

᾿Αλφαῖος 91 

ἅλων for ἅλως 121, 127 

ἀλώπηξ 130, 377 

ἅμα 328 

ἀμαρτωλός 158, 361 

ἀμέθυστος 370, 371 

ἀμήν 154 

ἀμήτωρ 276 

ἀμνός 135 

ἀμφί 294--ἀμφι- epds. 294 

ἀμφιάζω for -έννυμι 10, 214, 228—for 
-éfw 68, 228, 405 

ἀμφίβληστρον 294, 369 

ἄμφοδον 123, 295 

ἀμφότερος 182, 369 

ἄν for ἐάν 91 

ταν : acc. ending 3rd decl. 129 

-av for -ασι in perf. 221 

-ἂν (not -gv) 82, 84, 197 

ἀνά 295—dava- epds. 295 

ἀνάγαιον: Hellen. for -γειον 70, 76, 
296 

ἀναγινώσκω 295 

ἀνάθεμα and ἀνάθημα 57, 354 

dvakavifw, -6w 295, 394, 407 

ἁνακεφαλαιόω 296, 395 

ἀναλίσκω, -λόω 228, 296, 383 

ἀναλογία 296, 339 

ἀναπείθω 296 

ἀνάπειρος 72, 296 

ἀνάπεσε or -σαι 70 

ἀναπτύσσω 311 

ἀνασῖ for ἀνασείει 89 

ἀναστατόω 296, 394 

ἀναστρέφομαι, -στροφή 26 

ἁνδραποδιστής 280 365 

᾿Ανδρέας 119 

ἀνέκλειπτος (ἀνέγλιπτος) 108 

ἄνεμος 351 

ἀνένδεκτος 305 

ἀνεξίκακος 29] 

ἀνεξιχνίαστος 311, 371 

ἄνευ 328, 329 

ἀνήρ 136, 363, 433 

ἀνθομολογέομαι 297, 390 

ἀνθρωπάρεσκος 271 


INDEX OF WORDS AND FORMS. 


ἄνθρωπος 364, 433 

ἀνθύπατος 297, -evw 398 

ἀνοίγω 83, 188 f., 228, 294, 296 

-avés as suffix 359 

"Avravdpos 297 

ἀνταπόδομα 355 

ἀντί 296—avri- epds. 297 

ἀντιάζω 68 

ἄντικρυς 113, 328 

᾿Αντιοχεύς 350 

᾿Αντίπας 119, 297 

ἀντιπέρα 328 

ἀντίχριστος 297 

ἀντοφθαλμέω 297, 390, 392 

ἄνω or avd 58—dvw 296, 330 

ἀνώγεων 121 

ἀνωφελής 287 

ἀξιόω 397 

ἀόρατος 281, 371 

ἁπαλός 361 

ἀπαντάω 296, 384 

ἀπαντή 335— -ησις 335, 374—els ἀπάν- 
τησιν 463 

ἅπαξ 286, 376 

ἀπαράβατος 371 

ἀπάρτι, ἀπαρτί 299 

ἀπαρτίζω (-σμός) 299, 387 

ἅπας 285 

ἁπάτωρ 276 

ἀπαύγασμα 298, 854 

ἄπειρος 72 

ἀπέκδυσις (-εγδ-) 108, 373 

ἀπελεύθερος 299 

ἀπέναντι 328, 329, 465 

ἀπηλλαχώς 298 

ἀπλοῦς 285 

ἀπό 297—d7o- cpds. 297 ff.—Semitic 
uses 29, 31, 460 

ἀπὸ μιᾶς 28, 461 

ἀποδεκατόω, -evw 394, 399 

ἀπόδεκτος 58, 370. 

ἀπόδημος, -ἐω 192, 299, 311 

ἀπόδομα 355 

ἀποθνήσκω 84, 96, 240, 298 

ἀποκαθιστάνω : spelling 99 

ἀποκαθίστημι 10—double augm. 189 

ἀποκαραδοκία 274, 299 

ἀποκεφαλίζω 288, 291, 408 

ἀποκριθεὶς εἶπεν 453 

ἀπόκριμα (for -σι5) 355 

ἀποκτείνω 245 

ἀποκυέω or -ὕω 60, 246 

᾿Απολλώς 121 

ἀπολογέομαι 192, 299, 389 

ἀποστείλω or -στελῶ 70 

ἀποστοματίζω 299, 408 

ἀποσυνάγωγος 299 

ἤΛππιος, "Ard., Ampia,’ Add. 102, 109 

ἀπρόσιτος 324 

ἀπρόσκοπος 287 

ἀπροσωπολήμπτως 26 

ap changed to ep 67 


INDEX OF WORDS AND FORMS. 


apa, dpa, dpa 58 

ἄραφος 335, 371 

ἀργός 158, 287, 309, 335 

ἀργυροῦς, 120 

᾿Αρεοπαγίτης 82, 151, 277, 366 

dpeckela 57, 339 

ἀριθμός 351 

-dpiov as suffix 346 

ἀριστερός 369 

ἄριστον 271, 372 

ἄρκαλος 112 

ἄρκος 112 

ἄρκυς 112 

ἁρμόζω 410 

ἁρμός 350 

ἄρνας : stem 135 

ἀρνίον 135, 345 

ἀρξάμενος and ἤρξαντο: Aramaic use 
of 454 

-apos, -apa, -apos as suffixes, 363 

aprafw: mixed stems 229, 404 

ἀρραβών 101 f., 136, 153 

ἄρρην, ἄρσην 103 f. 

ἀρσενοκοίτης 367 

ἀρτέμων 136, 355 

ἄρτιος 336 

ἀρτοκόπισσα 349 

ἀρτόκρεας 269 

ἀρτύω 398 

apxe-, ἀρχι- epds. 70, 277, 290 

᾿Αρχέλαος 70, 289, 290 

-apxéw: cpds. in 388 

ἀρχηγός 71, 275, 277 

-dpxns and -apxos epds. 124, 271, 276 

ἀρχιερατικός 379 

ἀρχιιερεύς for ἀρχιερεύς 91 

ἄρχομαι : redundant use of 454 f. 

-as: nouns in 375 

τας: Mames in, Greek 119, Semitic 
146 

-doatin 2 8. pres. mid. 9f., 198 

᾿Ασάφ 109 

ἀσέλγεια 287 

᾿Ασιάρχης 124 

ἄσοφος 276 

ἀσπάζομαι 405 

ἀσσάριον 347 

ἄσσον 164 

ἀστεῖος 78, 338—dor. τῷ θεῷ 166, 443 

ἀστήρ 137, 368 

ἀσχήμων 287, -μονέω 391 

ἄσωτος 281 

ἀτενίζω 304 

ἄτερ, 328 

ἄτοπος 287 

av: sound of 43—changed to ἃ 87 

αὐθάδης 272 

αὐθαίρετος 272 

αὐθεντέω 278 

αὐθόπται 98 

αὐλέω and -ifouar 409 

αὐξάνω: augment 87 


519 


Αὐξιμητόρειος, Αὐξισπόρειος 291 

αὔξω 185, 229 

αὐτάρκης 272 

αὐτοκατάκριτος 212 

αὐτόματος 158, 272 

αὐτόπτης 98, 276, 367 

αὐτός 181 

αὐτός epds. 272 

αὑτοῦ (etc.) 87, 100, 180 

αὐτόφωρος 272 

αὐτόχειρ 283 

αὐχμηρός 87, 363 

ἀφαίρεμα 355 

apedpwv 136, 299, 356 

ἀφειδία 348 

ἀφελότης 287, 367 

ἀφελπίζξω (for ἀπ-) 98 

ἀφίημι 202-206, 241—ddets for ἀφίεις 
89, 202---ἤφιεν 189, 192, 202—ddes 
(-ere) as quasi-auxiliary 210 

apidw et sim. 98, 252 

ἄφιξις 298 

ἄφρων 134 

ἀφυπνόω 298, 394 

ἀφυστερέω 298, 393 

᾿Αχαία 84---Αχαϊκός for ᾿Αχαιϊκός 81 

ἀχρεῖος, -εόω, -ειἰόω 58, 81, 287, 338 

ἄχρι, -ἰς 113, 164, 329 

ἄψινθος, -θιον 123, 343 

-dw and -éw vbs.: mixed flexions 195 


B for a w sound 110 
βάθεα and βαθέα 60 
βαθμός for Attic βασμός 112, 351 


Baivw: simplex nearly extinct in 
Hellen. 11, 230—2 aor. imper. -βα, 
-βηθι 209 

Batov 342 


βάλλω: see 230—causative of πίπτῳ 
477 

Bapéw and βαρύνω 230, 387, 402 

BapoaB(B)as 102 

βασιλικός 378 

βασιλίσκος 380 

βασίλισσα 348 Ff. 

βαστάζω 10, 387, 405 

βάτος 123, 153 

βατταλογέω 68, 272, 389 

Bat =digamma. See F 

βδελυκλέων 290 

βδελύκτροπος 403 

βέβαιος 157, 337 

βεβαιώτρια 349 

βέβηλος 158, 360 

Βεεζεβούλ 105 

Βελίαρ 103 

βελόνη 357 

βέλτιον 164 

Βερνίκη 64, 92, 290 

Βηθσαϊδά 84 

βήριλλος, βήρυλλος, βηρύλλιον 79, 362 

βιάζομαι 400, -αστής 365 


520 


-βιβάζω 405 

βιβλίον (BvB-) 79, 343, -ἰίδιον 346, 
-dptov, -apld.ov, -ἰδάριον 347 

βιωτικός 379 

βλάπτω 403 

βλαστάω (-dvw) 214, 231, 382 

βλασφημέω 16, 290— -μος 272, 352 

βλέπειν ἀπό 31, 460 

BAnréov 224 

βόθυνος 358 

Boppas 92, 113, 119 

βουλευτήριον 343 

βούλομαι 231: augm. with ἡ 188— 
2 8. pr. ind. mid. -e for 7 97, 197 

βοῦς 127, 142, 333 

βουστροφηδόν 40 

βραδυπλοέω 284, 390 

βραδυτής 368 

βωμός 350 


vy as guttural nasal 41, 108 

γάγγραινα 348 f. 

γαζοφυλάκιον 272, 342 

Τάϊος not Tatos 50 

γάλα 133 

γαμέω, -lfw, -ioxw 231, 383, 409 

γαστήρ 137, 368 

γείτων 135 

γελάω 385 

γέλως 133 

γενέθλια 112, 364 

γένημα dist. from γέννημα 101 

γεννάω 356, 384 

Tev(v)noapér 102 

γεωργός 271, -cov 341 

γῇ 119 

γῆρας 140 

γίνομαι 108, 232—yéyova 220--- γέγονεν 
9--- γενόμενος meaning of 30—xal 
ἐγένετο 18, 425—~yiv. c. part. 452 

γινώσκω 108, 185, 232, 383—forms: 
γνοῖ 2 aor. subj. 83, γνοῦναι for 
γνῶναι 211 

γλῶσσα 348 

γλωσσόκομον 6, 58, 272 

γναφεύς 108, 349 

γνωρίζω 408 

γνῶσις 374 

γνώστης 366 

γόης 127 

Γολγοθά (for -γολθά) 105 

Γόμορρα 109 

-yovéw: epds. in 388 

γόνυ 133 

γονυπετής 271 

γραφεῖον 344 

γραφή 335 

γρηγορέω 220, 386 

γυμνιτεύω 72, 399 

γυναικεῖος 338 

γυνή 129f., 377 


INDEX OF WORDS AND FORMS. 


δ intrusive between ν and p 

Δαβίδ and Δαυείδ 86, 110 

δαιμονιάζομαι 406 

δαιμόνιον 341 

δάκρυον and δάκρυ 127, 141 

δάκτυλος 362 

Δαλματία (Δελμ-) 67 

δάμαλις 362 

δανίζω (for -εἰζω) 77 

δειγματίζξζω 408--- -ἰσμός 351 

δεινά, δεῖνα : accent 58—detva: flexion 
179 

δεῖπνον : gender 123 

δεισιδαίμων 291 

δέκα 107 ff.—dexadvo, δεκαμία 171— 
δεκαοκτώ 91, 171---δεκάπρωτος 276 

δεκανός 276 

δεκατόω 394 

δεξιολάβος 272, 334 

δεξιός 336 

δέρρις 104 

δεσμὰς δεσμάς, 270, 439 

δεσμός 121, 351 

δεσμωτήριον 342 

δευτεραῖος 176, 337 

δευτερόπρωτος 175, 269 

δέχομαι, Ion. dex- 108 

-Onuéw: epds. in 388 

Onunyopéw 273, 389 

δημιουργός 278 

δῆμος 351, 355 

δημοσίᾳ 84, 163 

δηνάριον 347 

δι- (=6Fi-) epds. 286 

διά 800---δια- epds. 301 ff.—7éurev 
διά 29, 402 ---διὰ χειρός 462 

διαζεύγνυμαι 301 

διαζώννυμαι 301 

διακατελέγχομαι 301 

διακονέω 192, 195, 303, 390 

διακρίνω 302 

διακωλύω 302 

διάλεκτος 303, 371 

διάνοια 303 

διανυκτερεύω 399 

διασαφέω 302, 391 

διασείω 301 

διάταγμα 355 

διαχειρίζω 302 

διδακτικός 379 

δίδραχμον 280, 286 

δίδυμος 352 

δίδωμι: forms after -w and -όω vbs. 
10, 202-206, διδοῖ 88----δοῖς, dot aor. 
subj. 83, 210---δῴη, δώῃ 83, 211— 
δῶναι for δοῦναι 88, 211---δώσωμεν 
75—dwon 218, 233 

διενθυμέομαι 302 

διέξοδος 303 

διετές 280 

διηνεκής 303 

δικαιόω 397 


INDEX OF WORDS AND FORMS. 


-d.ov as suffix 346 

διόπερ 303 

διοπετής 271 

διόρθωμα -σις 303 

Διόσκουροι and -κοροι 88, 271 
διότι 303 

διπλοῦς 120, 1δ0.---διπλότερος 166 
διχοτομέω 281, 392 

διψάω 195 ff., 233, 253 

δίψος : flexion and gender 125 
διώκτης 366 

δοκιμεῖον 78 

δοκιμή 352 

δοκίμιος 336 

δόμα 353-5 

Aopkas 375 

δότης 366 f. 

δράκων 356 

δύναμαι 234—augm. 188---δύνῃ 206 
duvaréw 390, 392 

δύνω 214, 284. ἔδυν 208 f. 

δύο δύο 270, 439 

δυσβάστακτος 276, 371 
δυσεντέριον, -ἰα 125, 287, 342 
δύσκολος 281 

δυσμή 306 

δυσνόητά τινα 9 

δωρέα, δωρεία 82 

δωρεάν 163 


εἰ sounds 40 ff.—vowel changes 68 ff. 
e for ἡ 73 

ἐάν for ἄν after ὅς, etc., 92 
ἑαυτόν 87, 180f. 
ἑβδομηκοντάκις ἑπτά 175 
ἕβδομος 353 

ἐβίων 239 

ἔγγυος 307 

ἐγγύς 164, 329 

ἔγειρε or -ραι 70 

ἐγκάθετος 386 

ἐγκακέω 307, 392 

ἐγκαλέω 305 

ἐγκαταλείπω 305 

ἔγκλιμα for -σις 355 
ἐγκοίμητρον 369 
ἐγκομβόομαι, 305 

ἐγκόπτω 306 

ἐγκρατής, -εια, -evouar 306 £., 400 
ἔγκυος 306 

ἔγχρισαι or -toar 60, 265 
ἔδαφος 364 

ἐδέετο for ἐδεῖτο 90, 199 
ἔδετο for ἔδοτο 212 
-εδύησαν or -ἐδυσαν 209 
-éfw: vbs. in 68 
ἐθελοθρησκία 290, 339 
ἐθέλω not Hellenistic 89, 188 
ἐθνάρχης 124 

ἔθνος 138 f. 

ἔθος 139 

ἐθύθη for ἐτύθη 112 


521 


el negandi 468 

εἰ μή for ἀλλά 468 

εἶ μήν 71 

-ela as suffix 339— -εἰα 348 

εἰδώλιον 342 

εἰδωλόθυτος 271 

εἰκῆ 84, 163 

εἰλικρινής, -ta (or -εια) 100, 273, 348 

εἰμί: flexion 201 ff.—middle forms 
10, 203—accent 203 

-εἶμι see 235 

-εινός as suffix 357 

-etov as suffix 342, 344 

τεῖος as suffix 338 

εἰπέ, εἰπόν : accent δῦ, 57 f., 209 

-elpnuat for -ἥρημαι 96, 226 

-elpw: vbs. in 402 

els 134, 169, 333, 439—els . 
els 438 

els (neither εἴσω nor és in Κοινή) 82, 
304—eic- cpds. 304—Semitic uses 
of 174, 462 

εἰσακούω 304 

εἰστρατιώτης (prothesis) 89 

elrev, ἔπειτεν 68 

ἐκ, ἐξ 308—éx- cpds. 309—Semitic 
use of ἐκ 460 

ἑκάτερος, ἕκαστος 182 

ἑκατονταετής 58, 286 

ἑκατόνταρχος 124 f. 

ἔκγονος sometimes éyy- 108 

ἐκδέχομαι 310 

ἔκδικος 311 

ἐκεῖνος 89, 178 

ἐκζητέω 310 

éxxalw 309 

ἐκλεκτός 58, 371 

ἐκλύομαι (ἐγλ-) 108 

ἐκνεύω 309 

ἐκνήφω 309 

ἔκπαλαι 311 

ἐκπειράζω 309 

ἐκπλήττω OF -σσω 107 

ἐκριζόω 310, 396 

ἐκτός 329 

ἐκτρέφω 310 

ἔκτρωμα 311, 354, 387 

ἐκφύῃ or -φυῇ 60, 264 

ἐκχέετε 265 

ἑκών 158, 330 

᾿Ελαιῶν or’ Eady 60, 152, 356 

ἐλάκησε or ἐλάᾶκησε 246 

ἐλᾶν for ἐλαίαν in papyri 91 

ἐλάσσων (or -ττὴ 107, 164 

ἐλαττόω 107, 396 

ἐλαύνω 235, 402 

ἐλαχιστότερος, -τατος 166 

ἐλεάω and -έω 195-8, 235, 385 

ἐλεεινός 92, 126, 357 

ἐλελύκεισαν 10, 221 

ἔλεος 126 

ἐλεύθερος 363 


3 3 Kal 


522 


ἐλθέ : accent 209 

ἑλίσσω and εἰλίσσω 236, 403 f. 

ἕλκος 126, 129 

ἕλκω : stem ἑλκυσ- in aor. and fut. 236 

ἐλλογάω (or -éw) 196, 198, 307, 385 

-edos, -é\7n as suffix 361 

ἐλπίς (ἐλ.) 98 

ἐλπιῶ for ἐλπίσω 236 

ἔλυσες 10 

ἑλῶ for αἱρήσω 226 

-eua for -nua: nouns in 73 

ἐμᾶτόν 87, 180 

ἐμβατεύω 306, 400 

ἐμβριμάομαι, -douac 198-201, 305, 384, 
397 


ἐμπαιγμόνη : ἐν ἐμπ. ἐμπαῖκται 28, 420 

ἐμπιπλάω 205, 254, 384 

ἐμπαρία,. ἐμπόριον 78, 342, -evouat 305, 
400 


ἔμπροσθεν 329, 331, 465 

ἐμπυρισμός 351 

ἐμφανής, -ἰζω 236, 307, 408 

ἐν 25, 304—éev- epds. 305—év or ἕν 60, 
439—Semitic use of 23, 25, 27, 463 
—intrusive ἐν common in Kow7 464 
—éy τῷ c. inf. 25, 450f., 464 

ἐναγκαλίζομαι 406 

ἐνάλιος 336 

ἔναντι 329, 465 

ἐναντίος 308 

ἐνάρχομαι 305 

évaros 101 

ἐνδέης 307 

ἐν δοξάζομαι 305 

ἐνδύνῳ 305 

ἐνδώμησις 73, 307, 374 

ἐνέδρα, -evw 125, 307, 399 

ἕνεκα, -xev, εἵνεκεν 67, 329 

ἐνενήκοντα 101, 168, 172 

ἐνεός (spelling) 101 

ἐνεργής, -εια 308, -éw 308, 388 

ἔνεστι 309 

ἐνέχω 306 

ἐνθάδε 308 

ἔνθεν 308 

ἐνθυμέομαι 308, 392 

ἔνι 306 

ἐνιαυτός 308, 372 

ἐνίστημι 306 

évvéa 101, 167 

ἔννοια 307 

ἕννυμι (obsolete) 228 

ἔννυχος 307 

ἐνορκίζω 305 

ἐνοχλέω, 189, 306 

ἐνταφιάζω 307, 405 

ἐντέλλομαι, ἔνταλμα, ἐντολή, 306 

ἐντεῦθεν 308 

ἐντός 330 

ἐντρέπομαι, 306, 387 

ἐντυγχάνω, ἔντευξις 306 

ἐντυλίσσω 404 


INDEX OF WORDS AND FORMS. 


ἐντυπόω 395 

ἐνύπνιον 291, 307, 341 

ἐνώπιον 15, 330, 465 

ἐνωτίζομαι 308, 408 

ἐξαγοράζομαι 309 

ἐξαίφνης, ἐξέφνης 69F., 81, 311 

ἐξαλιφθῆναι 227 

ἐξαπατάω 311 

ἐξάπινα 163, 311 

ἐξαποροῦμαι 310 

ἐξαυτῆς 311 

ἐξέραμα 311, 354 

ἔξεστι 309 

ἐξετάζω 310, 405 

ἐξηγέομαι 310 

ἐξίστημι 309 

ἐξισχύω 310 

ἐξολοθρευθήσεται 71 

ἐξορκίζω, -ιστής 309, 365 

ἐξουδενέω -bw, -θενέω -bw, -θενίζω 72, 
111, 310, 392, 396, 408 

ἔξω 330 

ἐξῶσαι for ἐκσῶσαι 108 

ἑορτή 92 

-eos: Att. gen. sing. 160 

ἐπαινέσω 226 

ἐπάναγκες or ἐπ᾽ ἀνάγκαις 69, 315 

ἐπάνω 315, 330 

ἐπάρχειος, -xla 157, 315 

ἐπαύριον 315 

᾿Επαφρᾶς 119, 314— -όδιτος 314 

ἐπέκεινα 315, 330 

ἐπενδύτης 277, 366, -δύω 313 

ἐπηρεάζω 314 

ἐπηρώθησαν, ἐπειρασθ-, ἐπειρώθ- 72 

ἐπί 311---ἐπι- cpds. 312 ff.— multipli- 
cative use of L73—émi τὸ αὐτό 1738, 
473 

ἐπίβλεψαι or -έψαι 60 

ἐπίγειος 157, 315, 338 

ἐπιγινώσκω 10, 312 

ἐπιδιγνύμεναι (for -decx-) 108 

ἐπιδιορθόω 313, 395 

ἐπιεικής, -εἰα (or -ia) 89, 314, 348 

ἐπιθανάτιος 315 

ἐπιθυμέω 192, 314 

ἐπιζητέω 312 

ἐπικέλλω 243, 294, 312 

ἐπικεφάλαιον 315 

ἐπικουρία 314 

ἐπίορκος, -έω 99, 314 

ἐπιούσιος 91, 313 

ἐπίσημος 314 

ἐπισκοπέω 258, 386 

ἐπίσταμαι 312 

ἐπίστασται for -acda 110 

ἐπίσταται for ἐφ- 99 

ἐπιστόλιον, -ἰδιον 346 

ἐπιστομίζω 315, 408 

ἐπισυνάξαι 10 

ἐπισφαλής 314 

ἐπιτήδειος 315, 338 


INDEX OF WORDS AND FORMS. 


ἐπιτιμάω 312 

ἐπιφανής, -eva 314 

ἐπιφώσκω 263 

ἐπιχειρέω, 192, 315, 392 

ἐπόπτης 314, 367 

ἐπουράνιος 157, 315 

-ep- changed to -ap- 65 

ἐραυνάω 86, 356, 384 

ἐργάζομαι 189, 237 

-epyéw: epds. in 388 

ἔρημος 68, 157, 352 

ἐριθεία 57, 339 

ἔρις 131 

ἔριφος 364, 387 

-Epos, -epa as suffixes 363 

ἔρρεεν 90 

ἐρρέθην 73, 102, 235 

ἔρίργιμμαι 102, 198, 257 

ἔρρωσθε 102, 257 

ἔρχεσθε or -σθαι 70, 200 

ἐρωτάω 385 

-εσαι in 2 5. pres. mid. 10, 198 

ἐσθής, ἔσθησις 133, 374 

ἐσθίω, ἔσθω 92, 238, 397 

ἔσοπτρον 82, 304, 369 

ἑσπερινός 358 

ἑσσόω: Ion. for Att. ἡττάω 107, 240, 
396 

ἑστάναι, ἑστώς 232: ἕστηκεν or ἔστηκεν 
(impf.) 100 

ἔσχατος 329 

ἔσω 330 

ἑταίροις or ἑτέροις 69 

ἑτερο- cpds. 284 

ἑτεροζυγέω 284, 3900 

ἕτερος 182, 329, 369, 437 

-éTns as suffix 366 

ἔτι 293 

᾿Ἑτοιμᾶς for ᾿Ελύμας 83 

ἕτοιμος 58, 157, 353 

éros 139 

ἐτροποφόρησεν or ἐτροφ- 109 

ev: sound of 43—ed with augm. 191 f. 

εὖ in comp. 281, 287 

εὐαγγέλιον 281, 341— -ἰστής 365 

εὔδιος 289 

εὐδοκέω 292, 392—evdox. ἐν 464 

εὐεργέτης 276 

εὐθυδρομέω, 282, 391 

εὐθύς 164—inferential εὖὐθ., Aram. use 
of 446 

εὐκόλως 441 

εὐλογέω 389— -ητός 371 

evoddra 191, 200, 238 

εὐπαρέδρος 320 

εὐπερίστατος 282, 321 

εὐποιία 83, 282 

εὑράμενος 213 

εὑρέ: accent 209 

εὕρεμα 73, 355 

εὑρίσκω 86f., 192, 238 

εὕροισαν 196, 211 


VOL. II, PART III.—35 


523 


-evs as suffix 349 

εὐτραπελία 282 

εὔφημος 287 

Εὐφράτης 281 

εὐχαριστέω 391 

evxetov 344 

-εύω : vbs. in 398 

εὐωδία 348 

εὐώνυμος 287 

ἔφαπαξ 315 

ἐφ᾽ αὕριον 98 

ἐφ᾽ ἑλπίδι 9, 98 £. 

"Edecivos 359 

ἐφ᾽ Eros 98 

ἐφευρετής 314 

ἐφιδεῖν 98 

ἐφίορκος, -έω 99, 312, 314 
ἐφίστασθαι for ἐπίστασθε 100 
ἐφνίδιος (aidy-) 70, 81 
ἐφόπτης (for ἐποπ-) 98 
᾿Εφρέμ, -ραίμ 50 

ἐφφαθά 102, 109 

ἔχεα 215 

ἐχθές 81 

ἔχιδνα 857 

ἐχυρός, -όω: Attic for ὀχ- 71 
ἕψεμα 355 

τέω : vbs. in 385 

ἑώρων 188 Ff. 

ἕως 330—€ws ὅτου 179—éws πότε 14 


F Digamma: origin and disappear- 
ance 39, 41—wrongly used to 
explain ἑλπίς 99—numerical sign 
(otherwise stigma 5) for six 167— 
preventing contraction 195—later 
represented by β 110. 

Fidns 272 

-ξαλής 408 

FaNicxw 228, 297 

Fapjv 135 

Fapyw 227 

ξεθόνη 357 

Εειδ- 234, 251 

Εεικ- 235, Felkw 314 

Fex-, Féxwv 280, 330 

βέμεμι 236 

Fer- 237, 247 

Féros 280 

Féros 279 

Fipis 375 

Fopdw 281 

-Fopés and cpds. 273, 363 

Fpt- 398 


€: sound of 45—in Kow7 106—sub- 
stituted for σ before voiced sounds 
106, 257—in transliteration 107 

ζαφθανεί 153 

¢B-, ζμ- in Kow7 and in NT 106 

Ζεύς 142 

ξέω 195, 239, 381 


524 


ζῆλος 126 

ζηλοῦσθε, -c Par 69, 200 
ζηλοῦτε : subj. 9, 75, 196 
ζήσωμεν 74 

(jw 195 ff., 239 

ζιζάνιον 344 

ζυγός : gender 123 
ζωγρέω 288, 392 

ζῴον 84, 343 

ζωπυρέω 284, 391 


7: origin 39f.—in Attic, Ionic, Kowy 
71—replaced by εἰ 71—7 extruded 
after οἱ 73—7j- preverb 188, 294— 
augment with ἡ 188—y 48, 118 

-7 as suffix 335 

-nyéw, -ηγός cpds. 68, 71, 275, 389 

-ηγορέω, -ἤγορος cpds. 389 

ἡδέως : comparison of 165 

ἡδονή 357 

ἠδυνάσθην and -ἤθην 234 

ἤθελον 188, 294 

ἥκω: flexion 221 

ἡλικός 379 

-ndos as suffix 361 

ἡμεῖς and ὑμεῖς : confusion of 73, 79 

ἡμέρα : phrases with 440 

ἡμιθανής 282 

ἡμιόλος 176 

ἥμισυς: spelling and declension 10, 
161, 176f. 

ἡμίωρον 176, 280— -ἰον 341 

ἦν : subj. of εἰμέ 113 

τὴν as suffix 356 

-nvos as suffix 359 

-ἥρημαι or -eipnuat 191, 226 

ἦρον, but ἦρα, ἤρθην 84, 226 

-npos as suffix 363 

Ἡρῴδης 84 

-ῆς: Semitic names in 146 

‘Hoatas 84 

ἥσσων 107, 165 

ἡττάομαι 8384— -nua 107, 354 

ηὐ- in augm. syllable 50, 86f., 191 

ἦχος 71, 125, 126 

ἠχώ 126, 140, 143 

-jw: vbs. in 68 


θ and τ interchanged 100, 110 
-6- stem 133 

θαμβέω 387, -os 126 

Bappéw, but θάρσει, -etre 103 ἔ 
θέατρον 369 

θεῖον 344 

θέλω : augm. 188 

θέμα 73 

θεμέλιος 122, -ἰον 341 

-@ev; quasi-ablat. suffix 164 
θεοδίδακτος 271 

θεός, θεά 125 

θεοστυγής 213 

θεραπεία 339 


INDEX OF WORDS AND FORMS. 


θεράπων 356 

θέρμα or -μη 118, 352 
-θετέω : epds. in 389 
Θευδᾶς 88, 91 

θεωρέω 363, 391 
θηνσαυρός : spelling 106 
θηρίον 343 

θλίψις : accent 57 
-θμός as suffix 35] 
-θνήσκω 84, 240, 383 
θρησκεία 339 
θριαμβεύω 400 

θρίξ 130, 333 

-Opos, -Opa, -θρον 364 
Θυάτειρα 128 
θυγάτηρ 136, 368 f. 
θυμιάω and θυμόω 385, 395 
θυμομαχέω 273 

Bupeds 347 

θυρίς 815 

θυρουρός 75 

θυρωρός 273, 363 


t: origin 39—: subscript 50—: avex- 
φώνητον 84 

Ly ἢ, Ἢν, εἰ, Ol, UV APproximating sounds 

t for ἡ 72 

cand e, cand o,: and οι, « and εἰ 76 

. sounds: two successive coalesce 89 

-t: irrational final 113 

τα as suffix 338 

-caxés as suffix 377 

᾿Ιαμβρῆς 103 

τ-ιανός as suffix 359 

tara, ἰᾶται 60 

-ἰδιον as suffix 346 

ἴδιος 181—idig 84, 163—xad’ ἰδίαν 98 

ἴδον for εἶδον 77, 190 

ἰδού: ** Hebraic”? use of 15, 23, 447— 
ἰδέ : accent 209 

ἱδρώς 133 

Ἵεράπολις 278 

ἱέρισσα 349 

ἱεροπρεπής 284 

᾿Ιεροσολύμα (‘lep-), 
1411. 

ἱερουργεῖν 91 

ἱερωσύνη : Ion. for Att. ἱερεωσύνη 91 

Ἴεσσαί 84 

-(fw: vbs. in 406 

-inue 201 ff., 241 

-ικός as suffix 377 

ἱλάσκομαι 283 

ἱλαστήριον 342 

ἔλεως 121 

τίλος as suffix 361 

ἱμάς 134 

ἱμάτιον 78, 341, 342 

ἱμείρομαι 76, 251 

τίμος : adjs. in 158, 353 


Ιερουσαλήμ 101, 


INDEX OF WORDS AND FORMS. 


iva for rel. pronoun 434 ff.—for temp. 
particle 469—c. fut. ind. 75 

ἵνα μὴ 470 

τινός as suffix 358— -wos, -ivos 359 

τίον as suffix 340 ff. 

Ιόὀπ(π)ὴη 102 

-tos as suffix 336 

ἱπποτετρόφηκεν 288 

τις: barytones in, flexion 131 

τις (-(60s): nouns in 131, 375 

ἴσασι 221 

toos: accent 58—Icos epds. 291 f. 

lore: imper. or indic. 221 f.—tore 
γινώσκοντες 22 

lornxew for εἱστήκειν 77 

ἰστήλη, ἰστοργή (prothetic ι) in certain 
districts of Κοινή 89 

ἵστημι : see 241 

τιστί: suffix 163 

-irns as suffix 366 

ἰχθύς 141, 333— -διον 346 

ἴχνος 356 

᾿Ιωάνης : spelling 102 

Ἰωσήχ 108 


kK: k and y—contiguous k and o, x 
and x: Semitic words with 108 

κἀγώ 84 

καθαίρω or καθαιρῶ 58: see 241 

κάθαλος 317 

καθαρίζω 67, 242, 406, 409 

καθ᾽ ἕτος 98 

κάθημαι 206 f. 

καθημερινός 158, 291, 317, 358 

κάθησθε or -ῆσθε θ0--καθῆσται for 
-σθαι 110 

καθ᾽ idiav 98 

καθίζω 242, 409 

καθολικός 317, 378 

καθώς 317 

καί spelt κά in some MSS 81—various 
syntactical uses of 34, 420 ff., 469 

καὶ ἐγένετο: constrns. with 8, 18, 
425 ff. 

Κάϊν 84 

καινοφωνίας (for κεν-) 69 

καίω not κάω 81, 242 

κακοήθεια 284 

κακοπαθία 78 

κακοῦργος 91, 271 

Kakouxéw 392 

κάλαμος 351 

καλέω 242, 387 

καλοδιδάσκαλος 278 

κάλφος for kap- 103 

κάμιλος for -ηλος 72 

καμμύω 92, 243, 398 

Kavavaios 109, 150 

καπηλεύω 399 

κᾷτα 84 

κατά 315—xara- cpds. 316 f. 

καταγγελεύς 350 


525 


κατάγλωττος 317 

κατάγνυμι 189, 226 

κατάδενδρος 317 

κατακληροδοτέω 391 

κατακρημνίζω 291, 317, 408 

καταντήσειν 216, 219 

κατείδωλος 317 

κατέναντι 329, 465 

κατενώπιον 330, 465 

κατεπέστην 316 

κατέχεεν : aor. 9] 

κατήγορος, -έω 192, 317— -ἰα 317 

κατήγωρ 127, 138, 317, 363 

κατήφεια 317, 326 

κατηχεῖν 71 

κατιόω 395 

κάτοξος 317 

καυθήσωμαι 74, 219 

καυσόω 242, 395 

καυστηριάζω 342, 405 

καυχᾶσαι 9, 198 

κεῖμαι 206, 243 

κεραία, Kepéa 69, 81 

κέραμος 851— -ἰον 343— -ικός 379 

κέρας 132— -άτιον 343 

κερβικάριον 346 

κερδαίνω 60, 74, 243, 401 

κερματιστής 366 

κεφαλαιόω 395 

κημόω 395 

κῆνσος 106, 155 

κηπουρός 273 

κηρίαις 72 

κηρίον 343 

κἢρυξ (κηρύξαι): accent 57— -é 
377— -σσω 403 

κιβωτός 372 

κιθαρῳδός 271 

κιθών : Ion. for χιτών 100 

Kitté, -ισσα 349 

κίνδυνος 358 

kivéw 244, 382 

κλαίω not κλάω 51, 244 

κλαυθμός 87, 351 

κλάω 185, 382 

κλείς 131 

Κλεόπας and Κλωπᾶς 88, 91, 109 

κλίβανος : Ion. and Dor. for κρίβ- 103 

κλίμα: accent 57, 354 

κλινίδιον 346 

κλύδων 136 

κνήθω 383 

κοινωνικός 378 

κολαφίζω 364, 407 

κολλυβιστής 366 

κολλύριον 78, 343, 363 

Κολοσσεύς 73, 350 

κοπάζω 387, 405 

komma 39, 168 

κόρος 153 

κοσμάριον 347 

κόσμιος : flexion 157 


130, 


526 


κόσμος 351 

KpaBarros 102 

κράνιον 342 

κράσπεδον 278 

κραταιός 337— -όω 397 
κρέας 140 

κρείσσων, κράτιστος 165 
κρεπάλη (κραι-) 81, 361 
Κρήσκης and Κλήμης : flexion 134 
κριθόπυρος 270 

κρίμα : accent 57, 354 
κρίνουσιν, κρινοῦσιν 60 
κριτήριον 342 ἴ. 

κρύβω 214, 245, 382 
κρυφῆ 84, 163 
-κτεν(ν)ω 214 

κτῆνος 139 

κτίσις 374 

κυκλεύω 400 

κυλίω 214, 246, 397 
-kuvéw 346, 382 
κυπάρισσος 349 
Κυρήνιος 72 

κυριακός 377 

κύων 134 £. 

κωμόπολις 276 

κῶμος 801] 

Κῶς 121 


41, 45—variations between ἃ and p 
103 

λαβέ, λαβοῦ 55, 57, 209 

λαθρα 84, 163 

λακέω, 246, 388 

λαλέω 388 

λαμβάνω: forms and flexion 103, 106, 
246 f. 

λαός 68, 71, 121 

λάρυξ : spelling 108 

Aacéa, -aia 81 

Λεββαῖος 86 

λεγεών, -ίων 76, 1386, 155 

λείπω and λιμπάνω 214, 247 

λειτουργός 96, 278— -éw and -la 278 

λέλυκες 10 

λέντιον 76 

Λευεί 86 

λευκαίνω 401 

λεφέλη ἴον νεφ- 103 

λέων 356 

ληνός: gender 123 

λῃστής 96 

λιβανωτός 123, 372 

AvBeprivos 359 

λιθοβολέω 392 

λίθος : gender 123 

λιμός 123, 350 

λογεία 339 

-λογέω : epds. in 398 

λογικός 378 

λόγιον 341 

λόγοι 9 


INDEX OF WORDS AND FORMS. 


τλογος : epds. in 284 f. 
λοιμός 350 

-λος, -λη as suffixes 360 
Λουκᾶς, Λούκιος, Λεύκιος 88 
λυμαίνομαι 401 

Λυσανίας 119, 291 
λυσιτελέω 291 

λύχνος 122, 856 


μ: intruded -uin λήμψομαι, ἐλήμφθην, 
τλημπτός 10, 246 

-u.- stems assimilated to -ν- stems 134 

μαγδωλοφύλαξ 414 - 

μαθητής 366— -τεύω 400— -rpia 349 

Maé@aios 102, 144 

μακράν 163, 332 

μακροχρόνιος 286 

μαλακός 377 

μᾶλλον, μάλιστα 165—elative use of 
μᾶλλον 108---μἄλλον μᾶλλον 270 

Μαμβρῆ 108 

μάμμη 118 

μάντις 374 

μαραίνω 401 

μαραναθά 154 

μαργαρίτης 366 

μάρτυς 137 ἴ.---μαρτύριον 341 

μαστὸς 110, 372 

-ματ- : stems in 132 

μάταιος : flexion 157 

μάχαιρα: flexion 118 

-μαχέω : epds. in 389 

μεγαλαῖος 338 

μέγας: flexion 160—comparison 165 
---μέγας μέγας 30, 270 

μεγιστάν 135 

μεθοδεία 318, 339 

μεθόριον 318, 341 

μεθύσκω 354, 383— -υσμα 354 

μείζων: flexion 113, 161]---μειζότερος 
166—welfous not contr. from -oves 
161 

μείξω, ἔμειξα, μεικτός 78 

μελανοσπαλάκισσα 849 

μέλας 185, 160— -awa 348 

μέλλω: augment 188 

μέλος : flexion, 139 

μεμψίμοιρος 289 

μέριμνα 356— -dw 384 

μεσημβρία 103, 278 

μεσογενής 276, 286 

μεσονύκτιον 341, 387—peoa- for μεσο- 
73 

Μεσοποταμία 291 

μέσος 336 

μεσότοιχος 276 

μεσουράνημα 278, 354 

μετά 317—pera- cpds. 318 f.—Semitic 
uses of 466 

μετακαλοῦμαι 318 

μεταμέλομαι 318 

μετανοέω 318 


INDEX OF WORDS AND FORMS. 


μεταξύ 317, 324, 330, 467 

μεταπέμπομαι 318 

μετέπειτα 293, 318 

μετέχω 318 

μετεωρίζομαι 318 

μετριοπαθέω 284 

μέτωπον 318 

μέχρι, -ἰς 113, 164, 317, 330 ἢ 

-μή as suffix 352 

μηδείς, undels 111, 170 

μηλωτή 372 

μήν 136 

μητήρ 136, 368 f. 

μητρολώας 68, 83, 85, 119, 274 

μιαίνω 401 

μίγμα : accent 57 

-μίγνυμι 249 

Μιθραῖον 342 

μιμνήσκω (or -yoKw) 84, 249, 383 

μισο- epds. 289 

Μιτυλήνη, -ἰνη 72----Μυτιλήνη, -υλίνη 79 

τὰμαι for -σμαι in pf. pass. 228, 249 

μνᾶ 19 

μογίλαλος or μογγί- 106 

μοιχαλίς 362, 375 

μολύνω 402 

-μός aS suffix 350, 355— -μος 351 f., 
355 

μοσχάριον 346 

μυέω 388 

μυλών 136, 356 

Μύρ(ρ)α 101, 128 

μυωπάζξω 290, 405 

μῶμος 351 

μωραίνω 401 £. 

μωρός 58—uwpé 152 

Muvojs 86 ἢ. 


vy: assimilation of ν in ἐν and σύν 104 
—insertion of » 106 

-v; movable -ν 113—‘ irrational” 
final ν 113, 129, 139, 161 

Ναζαρά (-ér, -ἐθ) 107, 112, 149 

Ναζαρηνός and Ναζωραῖος 107, 150 

Ναιμάν (Νεμάν), Ναΐν 84 

ναός 68, 71, 121 

Νάρκισσος 349 

ναυαγός 274— -éw 391 

ναύκληρος 274—vavkAnpoudxiuos 270 

ναῦς 142, 333 

veavias 119, 340 

νεανίσκος 380 

Νεάπολις 278 

veounvia 91, 279 

véos: does not contract 91 

νεόφυτος 279 

νεωκόρος 71, 121 

νεώς, νηός 8.υ. ναός 

νήθω 383 

νηλεής 126 

νήπιος 280— -τάζω 405 

Νηρεύς 350 


527 


νήριτος 351 

νῆστις 132, 162, 287, 374 
νηφάλιος 76, 157, 362 

vito 250 

νίκημα (for νίκη) 355 

νικο- cpds. 289 f. 

Νικομήδεια : scansion of 94 
νῖκος and νίκη 126, 381 

νίπτω 214, 250, 403— -τήρ 368 
νομοθέτης 366 

-vos, -vy as suffixes 356 
vooods (-la, -lov): Hellenistic form 92 
vooratw 405 

νοσφίζω 408 

vouvex7s 271— -εχῶς 163 
νοῦς 9f., 121, 127, 142 

-ντ- : stems in 134 

Νύμφαν, Νυμφᾶν 58, 71, 118 
νυχθήμερον 269, 283 

-νύω: vbs. in, for -νυμι 202 ff, 
νῶτος : gender 124 


&: digraph 41, 45—advbs. in -ξ 376— 
nouns in -£ 377, 380 

ξενοδοκέω 391 

ξέστης 155 

ξιφίδιον 346 

EvAd prov 346 

ξύρασθαι or -ἄσθαι 200, 250 — -do Ow 199 


οἱ sound 43—o and a, o and w 73— 
o(v) represents a w sound 110 

6- (see preverb ὠ) 294, 402 

ὀβολός from -ελός 7} 

ὄγδουν (for -oov) in papp. 91, 175 

ὁδηγός 68, 71— -ἐω 389 

ὀδμή, ὀσμή 112, 352 

ὁδοιπόρος 271 

ὁδόν 332, 459 

ὁδοποιέω 389 

ddovs: flexion 134 

ὀδύνη 294, 358 

ὀδύρομαι, -puds 294, 330, 350 

ὀθόνη 357— -dviov 78, 343 

olyw 294 

olda: flexion 10, 221 

οἴει 97 

οἰκεῖος 338 

oikerela 339 

-olkéw: epds. in 389 

οἰκήτωρ, -τήρ 365— -τήριον 342 

οἰκιακός 377 

οἰκοδεσπότης 217] 

οἰκοδομέω 250— -δομή 335 

οἰκουργός 214 

οἰκουρός 273 

οἰκτίρω 78, 250, 402 

οἰνάριον 347 

οἰνοπότης 367 

τοῖος as suffix 338 

ὀκέλλω 243, 294, 330, 403 

ὀκταήμερος 176, 286 


528 


ὀλίγος : cpds. of 284 

ὀλίγος for ὀλ- (with οὐχ) 98 

ὀλίγως 163 

-ὄλλυμι 251, 382 

ὀλοθρεύω, -τής 71, 365, 399 

ὁλόκαυτος 276 

ὅλος : epds. of 284 

ὄλυνθος 387 

ὀμείρομαι 76, 100, 251, 294 

ὅμιλος 335— -éw 386 

ὄμνυμι ἐν 464 

ὁμο- ορᾶβ. 284 

ὁμοθυμαδόν 104 

ὅμοιος 58, 157 

ὁμολογέω ἐν 29, 463 

ὁμοῦ 328 

ὀναρίδιον 346 

ὄνειδος 294 

ὄνομα 354 

ὁντιναδηποτοῦν 179 

ὄπιθεν, ὄπισθεν 311, 331 

ὀπίσω 331 

ὀπτάνομαι 214, 382 

ὀπώρα 279, 331 

ὁράω 189, 193, 251, 384 

dpyua: accent 58 

ὀρέων 91, 139 

ὀρθοτομέω 274, 390 

ὀρθρίζω 407 

ὀρθρινός 358 

ὁρκίζω 407 

ὁρμητήριον 343 

ὄρνιξ 108, 130, 133, and ὄρνις 133 

-opos as suffix 363 

-os as suffix 334 

-ogay: impf. and 2 aor. 9, 194, 209 

ὅσιος : flexion 157 

ὅσον ὅσον 270 

ὀστέων 9] 

ὅστις : Oblique cases rare in vernac. 
Kow7 179 

ὀστοῦν ; flexion 121 

-οσύνη as suffix 358 

ὀσφύς 141, 333 

ὅτι for rel. pron. 436---ὅτι recitativum 
469—for temp. particle 469 

ὀτρύνω 402 

ov: sound changes 87 

οὐ μή: force of, in NT 23 

οὐδείς and οὐθείς 111 

οὐθαμινός 359 

οὔκουν, οὐκοῦν 60 

-οῦν not -oty infin. 82, 197 

οὐράνος 157 

ods 133 

οὕτως 112 

οὐχ for οὐκ 98, 100 

ὄφελον 191, 252 

ὀφθαλμόδουλος 271 

ὀφρύς 141, 333 

ὀχυρός, ὀχύρωμα 71 

ὀψέ 331] 


INDEX OF WORDS AND FORMS. 


ὄψησθε 218, 251 
ὄψιμος 353 

ὄψιος 336 

ὀψώνιον 341 

-6w: vbs. in 393 ff. 


π᾿ and ¢ 109 

πάθνη : Ionic for φάτνη 100 

παιδίον 345 

παιδίσκη 380 

παλαιός 337 

πάλιν 118---πάλιν inferential, Aramaic 
use of 446 

πανδοχεύς, -ecov 108, 350 

πανήγυρις 279 

πανοικεί 163, 279 

πανοπλία 284 

πανοῦργος 271 

πανταχῇ 84 

πάντη 84, 163 

παντοκράτωρ 284, 365 

παρά 318—in epds. 319f.—subtractive 
use in numerals 173—in com- 
parison 467 

παραβάλλω 319 

παραβιάζομαι 319 

παραβολεύομαι 320, 400 

παραδειγματίζω 320 

παραθαλάσσιος 158, 320 

παραιτέομαι 319 

παρακαλέω 319 f. 

παρακύπτω 319 

παράλιος 157 £., 320, 336 

παραλυτικός 379 

παραμύθιον 341 

παραπλήσιον 331 

παρασκευάζω 319 

παραφρονέω 320 

παραχειμάζω, -acla 319, 340 

πάρδαλις 362 

παρεδρεύω 399 

map-es- epds. 320 

παρεκτός 331 

παρελάβοσαν 9, 209 

παρεμβάλλω 320— -βολή 307 

παρένεγκε or -και 70 

παρεπίδημος 26, 315, 320 

παρηκολουθηκότι ἄνωθεν 20 

παροιμία 320 

παρομοιάζω 320, 405 

παροτρύνω 320, 402 

παρουσία 320 

παροψίς 320, 375 

παρρησία (-ιάζομαι) 101, 192, 404 

παρρών 102 

ras: Semitic use of πᾶς οὐ 22, 433 f. 

πατήρ 136, 368 f. 

πατριάρχης 124, 271 

πατρίς 376 

πατρολώας 68, 83, 85, 119, 274 

πατροπαράδοτος 271 

| πατρώνισσα 349 


INDEX OF WORDS AND FORMS. 


πατρῷος 84, 338 

πεδινός 359 

πεζῆ 1085. πεζός 336 

πειθαρχέω 290, 388 

πειθοῖς or πιθοῖς 78 

πειθώ: flexion 143 

πεῖν for πιεῖν 89 ἔ. 

πεινάω 195 ff., 253 

πειράζω 387, 404 

πέμπειν διά 29, 462 

πένης 162 

πεποίθησις 374 

πέρας 132, πέραν 331 

περί 321—in cpds. 321 f. 

περιαιρέω 321 

περιάπτω 321 

περιεργάζομαι 321 

περιέχω 321 

περιίσταμαι 321 

περικεφαλαία 322 

περικρατής 322 

περικρύβω 245 

πέριξ 322, 376 

περιούσιος 322 

περιποιοῦμαι 521 

περιραίνω 250 

περισσεύω 192, 399 

περισσός, -ότερος 167, 192, 321f. 

περίψημα 322, 354 

πέρπερος, -evouat 270, 400 

πέρυσι 163, 279, 331 

πετεινός 357 

πηδάλιον 342 

πηλίκος 8, 24, 179, 379 

πηχύς 1401.---πήχεων 
91, 141] 

πιάζω 69, 254, 405 

πιέζω 69, 254, 311, 331 

πικραίνω 401 f. 

-πίμπλημι, -πίμπρημι: 
nasal 106 

πίμπρασθαι or -ᾶσθαι 60, 207, 2654 

πινακίς 375— -ἰδιον 346 

πιστεύειν εἰς 409. π, ἐν 464 

πίστις 374— -ικός 379 

πλανάω 384 

πλάνης 162 

πλάσσω 404 

πλέον 82-- πλέον πλέον 2 70---πλεονάζω 
82, 405 

πλεονέκτης 82, 274 

πλέοντα for πνε- 103 

mAnuvpns: spelling, 
274 


uncontracted 


insertion of 


101---πλήμμυρα 


πλήν 331 

πλήρης: flexion 162 
πληροφορέω 275, 390 
πλησίον 331 

πλοῦς 121, 127, 142 
πλοῦτος 127 

ποδήρης 275 

ποία (grass) 83 


529 


ποιέω: princ. pts. and orthography. 
255 £.— -ποιέω : cpds. in 389 

ποίμνιον 342, 346 

ποιός (ποιός): epds. in 386 

πόλεμος 352 

moNt for -u in Ionic 89f. 

πολιτάρχης 124, 289 

πολιτεύομαι 399 

πολλά: adverbial 446 

πολλάκις 164 

πολλαπλασίων 285 

πολυ- epds. 285 

πόμα for πῶμα 354 

πορεύομαι 400 

πόρνων or πορνῶν 60 

πόρρω 163, 323 

πορφυροῦς 347 

ποταμός 352 

ποταμοφόρητος 271 

ποταπός 112, 271, 375 

πότερος 181, 369 

Ποτίολοι 76 

morés and πότος 254, 370 

Ilovdns 134 

πούς 131, 333 

πραγματεύσασθαι or -σθε 70, 200 

πρᾶος 84 

πρασιαὶ πρασιαί 270, 439 

πράσσω 403 

πρεσβεία 339 

πρεσβευτής, -ύτης 86— -ῦτις 131, 376 

πρεσβυτέριον 341 

πρηνής 68, 323 

προ 322—in epds. 322 f. 

προαιρέομαι 322 

προαύλιον 323, 341 

πρόβατον 323, 372— -ἰον 345f. 

προδότης 323 

mpotuos 73, 353 

προίσταμαι 322 

προπάτωρ, 277 

προπωλήτρια 349 

πρός 323—in cpds. 324—=apud 467 

προσάββατον 323 

προσωγορεύω 399 

προσαίτης 365 

προσαχεῖν 71 

προσευχή 324-- προσευχῇ προσηύξατο 
28, 444 

προσεύχομαι 324 

προσήλυτος 237, 370 

πρόσκαιρος 291 

προσκεφάλαιον 324, 337 

προσκυνέω 246, 382— -ητής 365 

προσοχθίζω 408 

προστίθημι: Hebr. use of 445 

προσφάγιον 341 

πρόσφατος 323, 370 

προσωπολήμπτης (-πτέω, -Wla) 275, 366 

πρόσωπον 277, 324—in prep. phrases 
466 

πρόφασις 323 


530 


᾿προφητεία 339 

προφητεύω: augm. 189, 192, 256 
προφήτης 323, 366 

προφητικός 379 

προφῆτις 131, 376 

προχειρίζομαι 323 

πρύμνα 118 

πρωΐ 84. πρώιος 336 

πρωινός 73, 358 

πρῷρα 84 

πρῶτος 167, 323 

πρωτότοκος, πρωτοτόκος, -τόκια 58, 279 
πρώτως 163 

πτερύγιον 78, 343 

πτοέω 388 

Πτολεμαΐς for -αιΐς 81 

-πτω: Vbs. in 403 

πυγμή 352 

πῦρ 138 

πυρ(ρ)ός and πυρ(ρ)γάξζω 101, 405 
πωρόω 395 


p 41, 45—variations between ἃ and p 
3 


-pp- -p- 102, 192 f. 

-pp- -ρσ- 103 f. 

-pa nouns in: flexion 118 
-ραίνω 100, 256 

ῥάκος 152, 387-.--ῥακά 152 
ῥαντίζω 100, 256, 408 

pdoow 71, 256f., 403 

“Paxa8 and ‘Pad 109 

ῥέδη 81, 155 

pep-, pep- 100, 192, 256 
ῥηθῆναι 73, 235 

ῥήσσω ΤΙ, 102, 214, 257, 403 
ῥήτωρ: flexion 137 

purréw, ῥίπτω 257, 386 

ῥῖψαν : accent 57 

“PoBoau 109 

ῥοιζηδόν 164 

-pos, -pa as suffixes 362 

ῥύμη 352 

ῥυπόω, -αίνω, -apds, -apevoua 395, 400 ἢ, 


σ 45—final -s movable 112—stems in 
-o- 138, 381— -σσ- and -rr- 104, 107 

σάββατον 128, 153 

Σαλαμίς 128, 136 

σάλπιγξ 108, 3880— -πίζω 257, 409 

Σαμαρεῖτις 375 

Σαμοθρᾷκη 84 

Σαμψών 103 

σάν or σαμπῖὶ 39, 168 

-σαν in 3 pl. impf. 194—aor. 209 

σανδάλιον 343, 387 

Σάπφειρα 118, 145 

σάρδιον 124 

σαρδιόνυξ, σαρδόνυξ 90, 279 

σαρκικός, σάρκινος 378 

σαρόω 395 

σάτον 153 


INDEX OF WORDS AND FORMS. 


σαυτόν and αὑτόν 180 f. 
σβέννυμι 106, 257 

σελήνη 357— -τάζομαι 405 f. 
σεμίδαλις 362 

σεμνός : flexion 157 
Σερᾶπις : spelling 65 
σημειόω 396 

σήμερον 279 

σθενόω 396 

-σθωσαν in imper. 199 
-σία as suffix 340 
σιδηροῦς : flexion 120 
σικάριος 347 

Σίλας, Σιλουανός, Σιλβανός 109 f., 146 
σιμικίνθιον 72 

-σιμος as suffix 353 
Σίμων 146 

σίναπι 68, 140 

σινιάζω 405 

-σιος as suffix 337 

σιρικός 72, 378 

σιρός or σειρός 78 

-σις as suffix 355, 373 
σιτομέτριον 341 

σῖτος, 122, 372— οίον 343 
σκάνδαλον 361 

σκέπτομαι 214, 258, 403 
σκεῦος : flexion 139 
σκληροκαρδία 279---σκληροτράχηλος 285 
σκοπέω 258, 386 
σκορπίζω 387, 407 
σκότος : gender 127— -εἰνός 126, 357 
σκύβαλον 275 

σκυθρωπός 285, 364 
σκῦλον : accent 58 

-σμα as suffix 354 

-σμός as suffix 350 
Σολομῶν 146 

σουδάριον 347 

σπαράσσω 404 

σπάω 185, 258, 382 
σπεῖρα 117 

σπερμολόγος 275 

σπιλάς 360, 375 

σπίλος 57, 360— -όω 396 
σπυρίς : see σῴ- 

-σσω: vbs. in 403 

-oTa and στῆθι 209 
στάδιον 122 

στάμνος : gender 124 
-στάνῳ 214 

στατήρ 368 

σταυρίσκω 383— -pdw 396 
στείλω 214 

στείρα 118, 157 

στενάζω 405 

Στεφανᾶς 119 

στήκω 220, 259—or7jKere or -τε 73 
στηρίζω 259 

στιβάς (στοιβάς) 76, 375 
στοά: flexion 118 
στρατεία, -id: -la 78 


INDEX OF WORDS AND FORMS. 


στρατεύομαι 400 

στρατηγός 68, 71, 274 

στρατοπέδαρχος 124 

στρῆνος 127, 381— -τάω 385 

στῦλος : accent 58 

συγγενής 131, 138 f., 142, 162, 325— 
-εῦσι 138, 350 

ovyyevls 131, 375 

συγγνώμη 325 

συγχράομαι 325 

συκάμινος 153, 359 

συκομορέα 81 

συκοφαντέω 275 

συλλυπέομαι 325 

συμβαίνω 325, 427 

συμβιβάσει for LXX συμβιβᾷ 187 

συμβούλιον 341 

συμπόσιον 84]---συμπόσια 
270, 439 

συμφέρω 325 

σύν 324—in epds. 324 f.—Hebraism 
in Aquila’s use of, 17—assimilation 
of ν in συν- 105 

συναγωγή 325 

συναλλάσσω 325 

συναντή 335 

συνάξαι 10 

συναρμολογέω 272 

συνέδριον 325, 341 

συνείδησις 325, 374 

συνέκδημος 511 

συνελαύνω 325 

συνεργέω 192 

συνετίζω 407 

συνευδοκέω 325 

συνευωχοῦμαι 282 

συνέχεον 91, 265 

συνζητητής 365 

συνθλάω 240, 325, 382 

συνίημι 202 ff., 325—avvlovow, -otow 
60 

συνιστάνω 325 

συνομορέω 284, 386 

συντέλεια 325 

συντετρίῖφθαι not -τρίφθαι 57 

συνυπουργέω 389 

Συροφοινίκισσα 279, 349 

σύστασις 359 

συστατικός 325, 379 

σύστημα, -eua 57, 355 

σφ- for σπ- 109 

σφάγιον 336, 341 

σφυδρόν 103, 112 

σφυρίς 109, 375 

σχεδόν 164 

σῴζω 84, 260, 407 

σῶμα 355: phrases with 440 

σωπᾶν for σιω- 90 

σῶς : epds. of 285 

Σωσικόσμιος 291 

σωτήρ 151. 365— -τήριος 157, 336— 
-τήρια 336 


συμπόσια, 


531 


τ and @ interchanged 100, 110 

ταλαίπωρος 282 

ταλαντιαῖος 337 

ταμεῖον 89f., 344 

ταπεινός 357 

τάχειον (τάχιον) 164. 

τέθεικα, τέθηκα 72, 261 

τεθνάναι 240 

τείσω, ἔτεισα 78, 261 

τεῖχος : flexion 139 

τέκνον c. gen.: Semitic use of 27, 441 

τέλειος 82 

τελειωτής 365 

τελεσφορέω 275, 390 

τέλλω 261 

τελώνιον 342 

-τέος : verbals in 188, 224 

τέρας: flexion 132 

-Tepos as suffix 369 

τεσσαράκοντα 66 

τεσσαρακονταετής 38, 91, 286 

τέσσαρες : spelling 66—acc. 130, 170 

τεταρταῖος 176, 337 

τετρα- cpds. 286 

τετραάρχης 63, 91, 124, 276, 388 

τετράδιον 176, 343 

τέχνη 356 

τηλαυγής 283 

τηλικοῦτος 379 

-τήρ, -Twp, -της, -τής as suffixes 131f., 
365-369 

-τήριον as suffix 342 f. 

-τήριος as suffix 336 

-τικός as suffix 379 

τιμο- cpds. 289 f. 

τιμωρός 273 

τοιόσδε 178 

-τομέω : epds. in 390 

τοπάζιον 343 

-ros: verbals in 188, 224, 370—abl. 
suffix 329 

τοῦ c. inf. 448 ff. 

τράχηλος 361 

τρεισκαίδεκα 78 

τρι- epds. 286 

-Tpia as suffix 349 

τρίβος : gender 124 

Tpotro- and τροφοφορέω 390 

-Tpos, -Tpa, -Tpov as suffixes 368 

τροχός : accent 59 

τρύβλιον 342 

Τρῳάς 84 

-τωσαν in imper. 10, 195f. 


uv: itacism 78 f., 82—diphthongs with 
υ 85 ff.—vw not a diphthong 88f. 

ὕαλος (ὑάλινος) 67, 124 

tauwos 359 

ὕβρις 293 

ὑγ(ε)ῖα 89, 90 

ὑγιής 281—Att. acc. sg. ὑγιᾷ 139 

ὕδωρ 133 


532 


-via participles in: flexion 118, 159 
υἱός c. gen. : Semitic use 441 
ὑμεῖς and ἡμεῖς : confusion of 73, 79 
ὑμῶν αὐτῶν 181 

τύνω : vbs. in 402 
ὑπακούω 327 

ὕπανδρος 328 

ὕπατος 107 
ὑπεκπροφυγών 322 

ὑπέρ 326—cpds. of 326 
ὑπέρακμος 326, 352 
ὑπεράνω 326, 331 
ὑπερέκεινα 326, 331 
ὑπερήφανος 326 
ὑπερνικάω 326 

ὑπέρογκος 326 

ὑπερῷον, 84, 326, 338 
ὑπηρέτης 30, 328, 366 
ὑπό 327—epds. of 327f. 
ὑπόδικος 328 

ὑποζύγιον 328, 341 
ὑποκάτω 328 
ὑποκρίνομαι 327 
ὑπολαμπάς 328 
ὑπολήνιον 328, 342 
ὑποπόδιον 328, 342 
ὑπόστασις 328 

ὑπωπιάζω 75, 328, 405 
-upos as suffix 363 

τύρω : vbs. in 403 

ὃς 142, 333 

-ῦς for -vas ace. pl. 142 
ὕσσωπος 100, 124 
ὕστερος 167, 293 

ὑφαίνω 401 

ὑψηλός 327, 361 

ὕψιστος 165, 327 


g: sound 45—misplacement of as- 
pirate 98—doubled aspirate 102— 
¢ for π 109 

φάγομαι 238 ἢ 

φάγος : accent 59 

φάνῃ or -ἢ 60, 262 

φαντάζω 387, 405 

φαρμακεύς 349 

φαύσκω and φώσκω 263 

φελόνης, φαιλόνης 70, 81, 106, 155 

φερνάριον 347 

Φῆλιξ : accent 57 

-φημέω : epds. in 390 

φημίζω 387, 407 

φθινοπωρινός 279, 358 

φιάλη 67 

φιλάδελφος 290 

Φιλιππήσιος 337 

φιλο- epds. 289, 29] 

Φιλοκλέων 290 

φιλοπρωτεύω 399 

φιλοτιμέομαι 391, 393 

φλογίξω 408 

φλύαρος 158, 363 


INDEX OF WORDS AND FORMS. 


φοβηθῆναι ἀπό 29, 460 

φόβηθρον 110, 364, 369 

Φοινίκισσα, -vicoa 349 

φορέσωμεν 74 

φόρετρον 369 

-popéw: cpds. in 390 

φορτίον 342 

φραγέλλιον 103— -AAdw 396 

φρέαρ : stem and flexion 133 

φρεναπάτης 275, 367 

φρῆτος from φρέαρ in inscr. 91 

-ppovéw: cpds. in 391 

φρόνιμος 158, 353 

φρυάσσω 404 

Φύγελος 101 

φυλακτήριον 342, 365 

φυσιοῦσθε : subj. 9, 75, 200— -σθε or 
-σθαι 70 

φυτεία 339 

φωλεός 347 

φῶς : stem and flexion 133—accent 
133 

φωσφόρος 271, 289 


x and x 108—Semitic words with x 
108 

χαίρω : form of fut. tense 264 

χαλινός : flexion 122 

χαλκεύς 85, 349 

χαλκηδών 376 

χαλκίον 343 

χαλκολίβανος 280 

χαλκοῦς 121, 347 

Xavavala 109 

χαρακτήρ 368 

Xapav 102 

χάρις : flexion, 132—ydpw 331 

χεῖλος : flexion 139— -λέων 91, 139 

χείμαρρος (or -ppovs) 121, 271 

χείρ 137£., 333 

χειρόγραφον 27) 

χειροτονέω 275 

χέω 195, 215, 382 

χθών 134 

χιλίαρχος 124 

χιτών 100, 136 

χιών 134, 333 

χλιαρός 67, 363 

χορηγός 71, 275— -γέω 68, 71, 389 

χορτάζω 405 

χορτάρακον 270 

χοῦς : flexion 127, 142 

χρᾶσθαι or χρῆσθαι 200, 265 

χρεοφειλέτης 73, 367 

χρηζω 408 

χρηματίζω 265, 408 

χρήομαι 195, 265 

χρηστός 265, 370— -evouar 400 

χρίσμα 57 

Χριστιανός 59, 72, 360 

Χριστός 59, 371 


χρυσοδακτύλιος 285 


INDEX OF WORDS AND FORMS. 


χρυσοῦς 120f£., 347— -ἐων 91] 
χρώς 138 

-χύννω 191, 195, 214 f., 265 
χωρήσειν 216, 219 

χωρίον 342 

χωρίς 331 


Y: digraph 41—nouns in -y 364 
ψακάδισσα 349 

ψευδής : epds. of 280, 285 

ψῆφος 364 

ψιχίον 345 

ψύχος 57— -ψύχω 266 

ψωμίον 345 


w: history 39—sound 40, 44— 
approximation of o and w sounds 
73 ff.—confusion between w and ov 
75—w and a 76—w sound 191 

ὠ- preverb 243, 293, 330 


533 


-ὦδης : adjs. ending in 283, 376 

ὠδίν 135 

-whéw 189, 266 

ὠκεανός 293, 330 

-ωλός as suffix 361 

των as suffix 356—Semitic names in 
146 

ὠνέομαι : augm. and redupl. 189 

ᾧόν 84 

-wos as suffix 338 

wpvoua 294, 330, 387, 398 

-ws adverbial: extended use in Heilen- 
istic 163—in cpve. and superl. 163 

-ωσύνη : suffix 73, 358 

-wrepos in comparatives 73, 165 

-wTns as suffix 366 

ὠτίον 346 

-wrés as suffix 372 

ὠφελέω 294, 388— -ia 78, 348— -ἰμος 
158, 353 


(Ὁ) MopERN GREEK. 


ἀγγαρεμένος 67 

ddepre, ἀδερφέ 58 

ἄθρωπος ὅ2-- ἄθρωπος τῆς μπιστοσύνης 
440 

αἰσθάνομαι (pronounced estanome) 45 

ἀνάμεσα 64 

ἀνέβα, -βᾶτε 210 

ἀνεβοκατεβαίνω 270 

ἀπὸ μνιά 462 

ἄρκος 112 

ἄς 210 

αὐτός (aftds), (ἀγτός 87 


γεναῖκα τῶν γεναικῶν 443 
γέροντας 145 

γιὰ νὰ μή 470 
γυναικόπαιδα 270 


δέκα ἐννιᾶ, δεκανν,ά 17] 
δέκα ὀχτώ, δεκοχτώ 171 
δέσε, δέσετε 215 
δυῶ(νε) 170 


εἶδα 231 

εἶναι (ΞΞ ἐστί) 208, 306 
ἕνας 488 

ἑπτά (6{|4) 

ἔρημος 58 

ἐρπίδα 130 

ἕτοιμος δ8 

ἐφέτος 98 


ἥμαρτο 214 
ἥμισν (δυό᾽ μισυ, τρεῖς ἥμισυ) 178 


θεμελιώνω 395 
θεο- superl. formations 166 


καθημερνός 358 
κάλλιον 164 
κλέφτης 119, 145 


λιοντάρι 347 


μαχαιροπέρονα 270 
μέ (for μετά) 93 
μεγαλύτερος ἀπό 467 
μέρα (ἡμέρα) 180 
μεσάνυχτα 73 

μισός 178 

μωρός 58 


νά 470, 476 
ξολοθρεύω 71 


ὀκτώ (ochtd) 45 
ὅμοιος 58 


παπᾶς 145 
πεγάδι 43 
πέφτη 174 
πέφτω 477 
πιάνω 69, 254 
πιό, πλιό 165 
πίστι(ς) 130 
π(λ)εκμότερος 166 
ποῦ 435 
πρωτύτερος 166 


σαράντα 172 
σκλάβας τῆς σκλαβιᾶς 443 
σπέρνω 214 


534 


στάνω 202 


INDEX OF WORDS AND FORMS. 


στέκω 43, 96, 259 


στόν 275 


συνέβη, συνέβηκε : constructions with 
428 


τέσσερις, τεσσάρω(ν) 171, 175 


τί καλά 472 


(c) HEBREW AND ARAMAIC. 


Τριάδα 176 
τρίδιπλος 176 


φαγᾶς, fem. φαγοῦ 59 


φελόνι 155 


φθάνω (ftdno) 45 


χερότερος 166 
χιλιάδες 173 


[See pp. 143 ff., 152 ff., for Greek spelling and inflexion of Semitic words, 
also pp. 470-477 for retroversion of difficult readings into Aramaic or 


Hebrew. | 


ΝΞ, PIN 


Aramaic inferential 
conjunction 446 


am 
TDN [V2 


Ξε καὶ ἐγένετο 420 fi. 
453 | 


(Aram. 17) 198 439 m temp. use of 447 
pox 357 fs PP 3 23) 
πῃ) VOR ΣΝ 437 ‘a1 471 
j>&, 78 469 a an 439 
bon bax 443 $bn Hiph. of 455 
xox 468 Sx: Hiph. of 455 
ox negandi 468 on ov 439 
max 433 > 438 
we 434 b> 425, 433, 438 
ΔῊΝ 453 τ Ν᾽ Aram.=Gy 6 437 
ac. inf., temporal 427, "9513 476 
450, 485 5 predicative or re- 
nba 468 sultant 423, 462— 
RY} ‘ia 430 c. inf., gerundival 
(33) ya regular .Targumic 448—jussive 450— 
equiv. of Hebr. 53 introduces new 
475 subject 460 
xya Aram. vb. =seek, pray cvndxd added to adj. to give 
475 force of superlative 
xyix 12 441 442 
12 dina 467 nord 454 
τ sign of gen. in Aram. πὰρ 467 
431 pibyd a7 463 
1,72 Aram. indecl. rel. = woavn? 463 
Hebr. wx 434—= poxdp sip) 463 
ὅτι or ἵνα 435-7, 469 329 15, 465 
31 471 nxqp> 463 
x7 Aramaic = Hebr. j2 470 mon nin 443 
x7 temp. use in Aram. vn Aram. temporal conj. 
447 446 
ONT. . . WNT 438 jp 443, 460 
ΟΠ c. inf. 445 x10 7D 28, 461 
7378 170 444 ye 475 
ἡ consecutive 423, 425 pyo Aram. =Hebr. Ὁ} 475 
—in apodosis 422 3am 466 


INDEX OF WORDS AND FORMS. 


wv) 433 

DD Ny} 278 

yor 485 

1D 376 

by 439 

“3p Sy 466 

VaR}. . . ΜΝ 453 
Dy ton ayy 466 
xpap’s 380 

1p 470 


οι wip 443 
912 476 
nyy 475 
Gey 22 
x2 nby 29, 462 


τ auxil. vb. in Aram. 


455 
ww 477 
way 446 


Il. INDEX OF SUBJECTS. 


ee oe 


N—see Sinaiticus 

ἐξ in transliteration 108 f. 

A—see Alexandrinus 

Ablaut—see Vowel-gradation 

Accent 49, 51 f.—rules of accentuation 
53—recessive accent in vocatives 
54, 59—accent of 2 aor. imper. δῦ, 
57, 209—shortened penuit of nouns 
in -μα 57, 354—distinguishing 
words 57—recession of accent in 
proper names 59—in μήτηρ. θυγάτηρ 
368—accent of Latin and of Sem- 
itic words 59—differentiation by 
accent 60—in oblique cases of 
monosyllables 129—of 3rd decl. 
nouns in -6 131—barytones in -ts 
131—of contracted adjs. in -οῦς 156 
—of ἔστι 203—of nouns in -ἰον 345 
—of nouns in 77s 367 f.—in -rnp 368 
—of verbal adjs. in tos 371—of 
ΤΡ. epds. 392 

Accusative: pl. in -es 170—sg. in -ν 
140 ffi.—3rd decl. in -αν 129—advl. 
use 332, 444 f., 459 

Acts of Thomas 365, 478 

Adjectives: declension of 156—sim- 
plicia with 2 terminations 157-— 
cpds. with 3 terminations 158— 
comparison of 164 

Adverbs : formation 163—comparison 
of 164—advl. numerals 175— 
Semitic substitutes for 443 

** Aeolic ᾽" forms 139, 216 

Aeschylus 231, 275, 290, 292, 391 f.. 
398, 444. 566 also Index I (g) p. 512 

Aktionsart 263 

Alcman papyrus 56 

‘“ Alexandrian ”’ variants 63 

Alexandrinus, Codex 127. 129, 181, 
191, 198 f., 250, 258, 262, 428, 437, 
440, 442, 449, 465, 476. 

Allegro utterance 62, 92, 99, 250 

Alphabet 37—history of 38 ffi— 
numerical values 167 

Ammonius (Grammaticus i/ii A.D.) 57 

Analogy 91, 98f., 106, 118, 126, 171, 
198 £., 218, 252 1., 261, 272, 274 f., 
277, 282, 286, 308, 317, 337 £., 353, 


357 f., 363, 382 f£, 388 f., 393, 396, 
398, 400 f., 405 

Anarthrous nouns: qualitative force 
431 

Anatolia 109 

Ancyra inscription 437 

Anlaut 280, 303, 374 

Anthology, Greek, 122, 163, 292, 366, 
394—-see also Index I (g) p. 515 

Antiatticista 132 

Aorist : strong 186, 208 ff—weak 186, 
214 ff.—mixed flexions 208 ff., 215 
—aor. for pres. 458 

Apocalypse—see Revelation 

Apocopated forms 172 

Appian 457—see also Index I (g) p. 515 

Aquila (ii/A.D.): Greek version of 
O.T. 17, 198 

Aramaisms 14 ff., 416, 419 f., 430 f., 
434 ff., 43711, 441, 445 ff, 451, 
455 ff., 459, 461, 463, 467 ff., 471 ff. 
—see also Semitisms 

Aratus 263—see Index I (g) p. 513 

Arcadian dialect 310 

Aretaeus (Medicus ii/a.D.) 123, 365 

Aristophanes 7, 122, 246, 270, 275, 
287, 290, 300, 346, 433—see also 
Index I (9) p. 512 

Aristophanes of Byzantium (Gram- 
maticus iii/B.c.) 46, 48 

Aristotle 48, 97, 122, 125, 127, 163, 
170, 246, 280, 290, 335, 378 f., 381, 
384, 390 ff., 395, 401, 407—see also 
Index I (g) p. 513 

Arrian 288—see also Index I (g) p. 514 

Artemidorus 379, 399—see also Index 
1 (g) p. 515 

“ Artemisia ” papyrus 46 

Article: declension 117—unusual in- 
sertion 430—omission 430—év τῷ 
6. inf. 25, 450 f., 464—rod c. inf. 
448 fi. 

Aspiration 97 fi.—successive aspir- 
ates avoided 130, 262, 364, 369, 407 
—analogous aspiration 46—see 
Breathings, Interaspiration, Psilosis 

Assimilation of ε to o 71—of v tov, of 
«to v 79—of ν in ἐν and σύν 104 


536 


INDEX OF SUBJECTS. 


Attic declension, 121, 156---“ Attic” 
future 218—Attic dialect 3, 10, 43 f., 
58, 67 ff., 70 f., 77 ff., 84, 88, 95 ff., 
99, 1091. 1071 118f., 124f. 
136 f., 139 ff., 160, 170, 189 f., 191, 
196 f., 203, 209, 215, 218, 229, 299, 
239, 242, 244, 255, 258, 272, 295, 
303 £., 327 ff., 342 f., 350, 355, 355 f., 
362, 365, 368, 445 

Atticism 2, 6, 127, 132, 138, 141, 217, 
278 f., 313, 375 

Augment : εὖ or ηὐ 86— with ἡ 188— 
double augm. 189—syliabic for 
temporal 189—dropped augm. 190 
—temp. augm. for reduplication 191 
—initial diphthongs 191—augm. 
in cpds. 192—vbs in p 193 

Avroman in Media : parchments from 
13 


B—see Vaticanus 

Babylonian Talmud 451 

Bacchylides papyrus 46, 49, 56 

Back-formations 68, 220, 228, 230 f., 
238, 241, 250 f., 335, 338, 356, 358, 
363, 382, 384, 386, 394, 402, 404 f. 

Bahu-vrihi (B.V.): Possessive Cpds. 
283, 393 

Barnabas, Epistle of 129—see also 
Index I (d) p. 503 

Barytones 118, 131, 142, 334, 367 

Belloc, Hiliare: bilingualism of 479 

Benedictus : Hebraic style of 482 

Bezae, Codex 31, 50 f., 56, 69, 75, 79, 
98, 100, 102 f., 108, 110, 136, 145, 
171, 194, 196, 208 f., 216, 240, 245, 
262, 210, 301, 322, 348, 356, 364, 
420, 423f., 480, 432, 437, 441, 
475 

Bilingualism in Galilee and Jerusalem 
12 f 


Boeotian 42 f., 77, 80, 82, 88, 95, 104, 
107 f., 174, 304 

Breathings 40, 45, 49—in MSS 97 f. 

Bunyan, John: Biblical language of 
8, 478 


Cappadocian—see Pontic 

Cardinals 169—Semitic use of 439 

Carpathos 107 

Cases: with prepositions 292—see 
under the several cases 

Casus pendens : followed by resump- 
tive pronoun 423 

Cato 360 

Causative force in composition of ava 
295— of κατά 316 

Chalcedon 376 

Charax, Joannes (vi/A.D.) 58 

Chios 107 

Chrysostom, 74, 282, 388, 443 

Cicero 360, 390 


537 


Clement of Alexandria 44]—Atticism: 
of 2, 6—uses γνώῃ subj. 211 

Clement of Rome—see Index I (d) 

Cleonae inscription 111 

Cobet, C. G. 122, 189—see p. xvii 

Combinations written as one word 63 

Comedy : Old 444—Middle 385—New 
317, 381 

Common Greek 1, 4, 11, 13, 17, 62, 
94. f., 130, 133—see also Hellenistic 
Greek, Kowy passim 

Comparative particle after positive 
adj. or vb. 442 

Comparison of adjs. and advbs. 164— 
double comparison 166—Semitic 
substitutes for comparative 440 

Composita 293-328 

Compound verbs: no evidence of 
Greek culture 11—double epds. a 
tendency in Hellenistic 389 

Compounds: Co-ordinatives (Dvand- 
va) 269—Iteratives 270—Depen- 
dent Determinatives (T.P.) 271—-5— 
Descriptive Determinatives (K.D.) 
276 ff.—Possessive (B.V.) 283 ff.— 
Verbal 288 fi.—Based on Prepo- 
sitional Phrases 291 f.—Preposi- 
tional Cpds. 294-328 

Conative force 311 

Confusion of Contracta 194 ff. 

Conglutinates : with -cov 346—in -θμός 
353—with -αδ-, -.d- 376 

Conjugation and tense stems of vbs. 
183—-conjug. classes of vbs. 184, 

81 fi 


Consonantal dissimilation 103, 106, 396 

Consonants : single and double 101— 
intrusive 103 

Construct state (Semitic) 25, 440 

Contract verbs 193—201—assimilation 
in 75, 195 ff. 

Contraction 89—reversion to 
contracted forms 90—contracted 
nouns 119, 127, 138—adjectives 
120—of gen. plur. 139—contracted 
fut. of -ifw vbs. 218 

Co-ordination of participle or infin. 
with finite verb 428 

Coptic influence 66, 87 

Crasis 55, 63 

Cretan dialect 296, 304, 329, 333—see 
Gortyn 

Cyril of Alexandria (iv/v A.D.) 74 


un- 


D—see Bezae 

Dative: Hebraic use of cognate dat. 
443—dat. of agent 459—after 
évoxos 460 

De-aspiration—see Psilosis 

Declensions: First 117—Second 120 
—Third 128—of Semitic names 143 
—of non-Semitic names 150 


538 INDEX OF 
Definition: gen. of 25, 27 f., 440, 485 
Delphic dialect 86, 194 

Demosthenes 343—see Index I (9) 


p. 513 

Dentals 110—dental stems 131— 
suffixes 364 

Dependent Determinatives (T.P. 
cpds.) 271—see Tatpurusa 
Descriptive Determinatives (K.D. 


epds.) 276—see Kharma-dharaya 

Deteriorative force of suffix -ἰον 344— 
of suffix -ίσκος, -ἰσκη 380 

Diaeresis 50 

Dialects in Kow7 95, 125 f., 131, 139, 
194, 196, 203, 210 

Dialects of ancient Greek 66 ff., 71, 77, 
119, 139, 170, 296, 304, 329f., 365— 
see under Attic, Doric, Ionic, etc. 

Dialects : pronunciation 41 ff. 

Didymus of Alexandria (iv/A.p.) 74 

Differentiation by accent 60 

Digamma 39, 99, 110, 133, 135, 158 f., 
167, 185, 1891., 195, 215, 227f., 
263, 272 ff., 280 f., 285 ff., 294, 296, 
300, 307, 314, 318, 330, 334, 338, 
342, 344, 348, 355, 357f., 360, 362 f., 
372, 382, 385 ff., 390, 398, 406, 408 
—see Index IT under F 

Diminutives in -ἰον 344 ff.—in -ἰς 375 
—in -ickos, -ἰσκη 380 

Diphthongs 44—with « 80—with υ 85 

Dio Cassius (ii/iii A.D.) 151, 391 

Dio Chrysostom (i/ii A.D.) 399 

Diodorus Siculus (i/B.c.) 193, 200 f., 
379, 390, 392 f. 

Dionysius of Halicarnassus 389—see 
also Index I (g) p. 514 

Dioscorides (Medicus i/A.p.) 407 

Directive sense in composition of ἐπί 
312, 314—of πρός 323 f. 

Dissimilation of vowels 79—of con- 
sonants 103, 106, 274. 364—of 
aspirates 130, 217, 262, 369—“ pro- 
hibitive dissimilation ” 401 

Distributives 175, 270, 437-440 

Dogma cause of solecism 154 

Doric 68 f., 70, 86, 98, 108, 119, 123, 
130, 178, 238, 254, 281, 303, 329, 
331, 367 

Double comparative and superlative 
166 

Dvandva (Co-ordinative Compounds) 
269 

Dvigu 276 


Egyptian inscriptions 87, 170, 330— 
Κοινή 170 

Elative 164 ff., 442 

Elian dialect 71 

Elision 61 f. 

Enclitics 54, 180, 203 

Epexegetic Infinitive 448 


SUBJECTS. 


Ephesians : Epistle to 22, 419 

Ephesus ; inscriptions from 87, 330 

Epic language 58, 329, 407 

Epicharmus (Comicus v/B.c.) 305 

Epictetus 317, 374, 392, 395, 410, 437 
—see also Index I (g) p. 514 

Kpiphanius (Bp. of Salamis iv/a.p.) 
74, 439 

Epsilon : origin of name 38 

Erasmus 154, 335 

Erythrae 330 

Etacism 95 

Ethnica 150, 350, 359, 366, 375, 378 

Etymon 112 

Euphemism 288 

Eupolis (Comicus v./B.c.) 407 

Euripides 52, 210, 292, 378—see also 
Index I (g) p. 512 

Eusebius of Caesarea (iv/4.D.) 126 

Eustathius, Abp. of Thessalonica 
(Grammaticus xii/A.D.) 390 

Eustathius, or EKumathius, Macrem- 
bolites (xii/A.D.) 365 

Ezra-Apocalypse (or 4 Ezra) 455, 480 
—see Index I (c) under 4 Ezra 


Faded diminutives 345, 380 

Fayyum 147—see also Index I (f) 
under P Fay, p. 507 

Final letters movable 112 f. 

Formative suffixes 333 

Fractions 174, 176 

Freer—see Washington Codex 

Frequentative verbs 385 f., 405 

Future tense 187, 218 f.—‘‘ Third 
Future’ 187—future infin. obsol- 
escent 216 f., 219—“‘ Attic ”’ future 
218, 236—fut. subj. imaginary 218 
—fut. opt. obsolete 219—future for 
imperative 458 


Geminus 137 

Gender: fluctuations of 
proper names, 152 

Genitive plural in -éwy or -ῶν 139—- 
“Doric”? gen. 119—‘* Hebraic” 
gen. 27f., 440—gen. of agent or 
cause 459—gen. of object 440 

Gentilic names 144, 150, 337 

Gerundive in -réos 188, 224 

Gortyn Law (Crete v/B.c.) 515 

Gutturals 108—guttural stems 130— 
and dental stems: confusion of 
flexions 229, 257, 404, 407—-suffixes 
376 


121 ff.—of 


Haplology : syllabic 68, 336, 365 

Harmonisation of Gospel texts 203 

Hebraisms 14-34, 416, 421f., 425, 
429f., 434, 438f., 440f, 442 £, 
444 f., 446, 451, 459f., 462, 464, 
468, 47]- 566 also Semitisms 


INDEX OF 


Hebrews, Epistle to: Greek style of 
6, 11, 24f., 219, 221, 467 

Hecataeus Milesius (vi/v B.c.) 407 

Hegelochus, story of 52 

Hellenistic flexions 10 

Hellenistic Greek: pronunciation of 
42 f.—accentuation 56 f.—see Κοινή 

Herculaneum 112 f. 

Hermas 129, 166, 192, 195, 199—see 
Index I (d) p. 503 

Herodian (grammarian ii/A.D.) 119, 
140 f. ᾿ 

Herodotus 103, 140, 151, 202, 204, 231, 
295. 304, 310, 338, 378; 381, 386; 
389, 417 f., 472—-see also Index I(q) 
p. 512 

Hesiod 209, 286, 295, 338—see also 
Index I (g) p. 512 

Hesychius (Lexicographus 
297, 306, 396 

Heteroclisis 124, 128, 132 ff., 136, 138, 
141 f., 350 

Hiatus 62, 91 

Hippocrates (Medicus v/s.c.) 109, 
170, 189, 191, 201, 209, 231, 250, 
261, 370, 393, 405, 408—see Index 
I (g) p. 512 

Hippolytus (ii/a.D.) 335 

Homer 131, 149, 234, 238, 263, 273, 
292 f., 295 f., 303 f., 331, 337 £., 345, 
350 £., 353, 359, 367, 369, 386, 391 f., 
397, 405 f., 408, 429—see also Index 
I (9) p. 512 

Hyperbaton 424 

Hypocoristic nouns 344 

Hypotactic force of καί 420 


v/A.D.) 


Imperative flexions 195-22: 
Imperfect tense: flexions 194-206— 
rendering Aramaic participle 457 
Impersonal use of 3rd plur. act. 30, 
447 

Indeclinable use of πλήρης 162—of 
λέγων (λέγοντες) 454—Indeclinables 
in -α 148 

Infinitive : Semitic uses of 448 fi.— 
Co-ordination of infin. with finite 
vb 430--ἐν τῷ c. inf. 25, 450 f., 464 
—-rod c. inf. 448 ff. 

Intensive force of ἐξ in composition 
309 

Interaspiration 98 

Interchange of n- and r- 401 

Intransitive verb used in place of 
normal passive 448 

Intrusive consonants 103, 106, 137, 
247, 366 

Inverse attraction of relative 424 

Ionic elements in Κοινή 4, 43, 66 ἢ, 
70 ff., 88, 91 f., 95, 99 f., 104, 108, 
118 f., 121, 125, 140, 174, 187, 195, 
198 f., 202, 204, 221, 231 f., 234, 


VOL. II. PART III.—36 


SUBJECTS. 539 


236, 275, 295, 303, 329 f., 355, 365, 
374, 381, 387, 398, 407, 448 

Iota subscript 50, 96, 115, 197, 211, 
260 

Irenaeus 25] 

Isaeus (iv/B.C.) 378 

Isaiah 17—see also Index I (ὁ) p. 501 

Isidore of Pelusium (v/a.D.) 279 

Isoerates 343, 378—-see also Index I (4) 
p- 513 

Itacism 65 ff., 72, 74, 76 f., 80, 83, 88, 
132, 190, 200, 205, 247, 339, 410— 
—and Syntax 74 f., 96 

Iterative Compounds 270—iterative 
force of present 265, 386 


James: Greek style of 10, 15, 26f., 
222, 

Jerome (iv/v A.D.) 313 

John: Greek style of 13, 31 f., 162, 
483 

Josephus 86, 119, 143, 147, 152, 170 f., 
175, 190, 192, 356, 365, 375, 390, 
408, 480—see also Index I (g) p. 514 

Jude: Greek style of 27 

Justin Martyr 202, 365, 390—see also 
Index I (d) p. 503 


Juvenal 279 


Kalidasa 268 

Karma-dharaya (K.D.) Compounds, or 
Descriptive Determinatives 276, 
367 

Καθαρεύουσα or Literary MGr. 465 

Κατήχησις. primitive 481 


Kidron: Greek popular etymology 
149 
Klaros: inscriptions from 274 


Kow7: vowel system in 64-95, 139— 
rule of contraction 89—earlier and 
later Κοινή 118, 122, 127f., 130, 198 
—FEastern Kow7 94, 121—Achaean- 
Dorian 170—Egyptian 170 

Kretschmer’s Law 56, 64, 92, 290 


Labials 109—labial stems 131—labial 
suffixes 364 

Laconian dialect 45, 353, 451 

Language conditions of Palestine 12 f. 

Latin: names 134, 155—loan words 
103, 106, 128, 155, 347 

Latinism 103 

Lesbian dialect 119, 139, 318, 352, 357 

Lincoln, Abraham : Biblical language 
of 478 

Lingua cottidiana in LXX and NT 2 
—lingua franca 1, 12, 413 

Liquids 41, 45, 103—liquid stems 136 
—liquid suffixes 360 

Loan words 127 f., 375, 398—see also 
Latin 

Locative terminations: accent 53 


540 INDEX OF 


Logistic : alphabet in Gr. logistic 169 

Lucian 120, 135, 279, 365, 379—~see 
also Index I (g) p. 515 

Luke: style of 3, 7, 18-21, 138, 162, 

216, 219, 222, 300, 357, 427, 432, 
447f., 451 f., 455, 473, 481, 483— 
Antiochene or Philippian ὃ 19— 
unity of Lucan writings 7 

Lycian inscription 202 

Lysias 189—see also Index I (9) p. 513 


Macedonian dialect 102, 290, 307, 320, 
348 f. 

Magdola 415 

Magnesia, 77, 80, 82f., 90, 94, 125, 
139, 181, 195—see also Abbrevia- 
tions, p. xxii, under Nachmanson 

Magnificat : Hebraic style of 417— 
Harnack’s judgment 482 

Manuscripts of NT: orthography 
tested 88, 90, 92, 111—see Alexan- 
drinus, Bezae, Sinaiticus, Vaticanus, 
Washington 

Marcion (ii/A.D.) 74 

Marcus Aurelius 400, 410—see Index 
I (g) p. 515 

Mark : style of, 11, 13, 28, 30, 446, 481 

Matthew : style of 10, 29, 481 

Medieval Greek 425, 431, 433 f., 435, 
438 f., 442, 446, 451 f., 461 ff., 465, 
467, 469, 476f.—see Abbrevia- 
tions, p. xxi, under McKinlay 

Megarian dialect 331 

Menander 61, 122—see also Index 1 (σ) 
p. 514 

Metaplasmus 119, 124, 147, 151 

Metathesis of aspiration 100, 108, 110 
—of quantity 121—of consonants 
106, 155 

Middle voice 197 ff., 205 ff., 223 ff.— 
middle forms of εἰμί 203 

Misplacement of h 98 

Mistranslations of Semitic words and 
phrases 470 ff. 

Mixture: of declensions 146, 151 (see 
Metaplasmus)—of conjugations 75, 
194 ff., 202 ff., 265 

Modern Greek 5, 14, 42 f., 51, 55 f., 59, 
63 ff., 66 f., 76, 86, 92, 95, 101, 104, 
107, 1135: 190: 155. τοῦ: 69} 
170, 172 £., 175, 177, 183, 192, 194 f., 
198, 202 f., 209f., 214f., 253, 316, 
318, 346 f., 353, 379, 387, 417, 420 ff. 
425, 428 f., 431, 435, 438 ff, 442, 
446, 448, 452, 456, 461 f., 465, 467, 
469 ff., 472, 476 f.—see also Index 
II (6) p. 533; Abbreviations, pp. 
xxi, xxiv, s.vv. McKinlay, Thumb 

Moeris (Grammaticus ii/A.D.) 57, 67, 
80: 119. 195 155. 132, 189592598! 
906. 217 f., 221, 407—see Abbrevia- 
tions, p. XXi 


Ϊ 


SUBJECTS. 


Mommsen +79 

Movable letters -s, -v, τι 112 

Multiplicative numeral adjs. 
multipl. use of ἐπί 173 

Mutes 40, 44 

Mysteries : ἐμβατεύω a technical term 
of 274 


176— 


Nasals 41, 45, 103—nasal stems 134 ff. 

—mnasal insertions in verbal stems 
184 

—suffixes 350 

Neuter nouns in -ἰον 340 

Nomina actionis 289, 334, 341 f., 355, 
373 f., 392 

Nomina agentis 289, 334, 337, 341 f., 
361, 364 ff., 368 f., 377, 385, 391 f, 
402 

Nominativus pendens 423 f., 447 

Non-assimilation of ν 105 

Nouns: declensions 117—word-com- 
position 268—suffixes 332 

Numerals 167—numerical signs 168 
—arrangement of 172—abstract 
numerals 176—multiplicative 176 
—‘‘ Semitic "ἢ use of 439 f. 

Nymphodorus (Historicus 3.8.0.) 317 


Onkelos: Targum of 453, 460 

Onomatopoeia 388, 410 

Optative 6 f., 53, 196, 200, 204, 207, 
911; 2032S tes 219 

Orators: the Greek 295, 399 

Ordinals 167, 173, 372 

Origen 175, 313 

Orthography 51 fi. 

Orthotone words 54 

Oxytone 53, 57, 59, 209, 328, 334, 359, 
367 


Palestine : language conditions of 12 f. 

Pallis: Modern Greek version of NT 
425, 429, 448, 470 

Pamphylian dialect 323 

Paragraphs 47 

Parallelism 417 f. 

Parataxis, 414, 420 fi. 

Paroxytone 53, 59, 371 

Participle : Semitic uses of 451—co- 
ordination of participle with finite 
vb 34, 428 ff.—redundant use of 452 

Patronymics 356, 375 

Paul: Greek style of 2, 8 f., 21, 162, 
221, 326 

Pausanias (ii/4.D.) 151 

Pentateuch: Greek of 
Index I (0) p. 501 

Penult shortened in many nouns in 
-μα 57, 354 


17 f.—see 


INDEX OF 


Perfect tense 187, 220 ff. 

Perfectivising force in composition of 
ava 296, ἀπό 297, 393, διά 301, ἐξ 
308 ff., κατά 298, 316, σύν 324 

Pergamum 67f., 72, 74, 77, 80, 82 f., 
90, 94, 99, 104, π|9 90 151, 171, 
181, 191, 199, 200, 206——see Ab- 
breviations, p. xxiii, under Schwyzer 

Periphrastic tenses 224, 239, 451 

Perispomenon 53 f., 117, 119, 146, 156 

Peter: First Ep. of 10, 13, 25 f.— 
Second Ep. of 5, 27 f., 420 

Philo 86, 292, 339, 365, 379, 391, 405 
—see also Index I (g) p. 514 

Philodemus (Philosophus i/B.c.) 288, 
390 

Philostratus 94—see also Index I (0) 
p- 515 

Phocis 194 

Phonology and writing 37-114 

Photius (Lexicographus ix/aA.D.) 55, 
131, 390 

Phrygian influence 89, 102—Phrygian 
Kow7 323—inscription 126 

Phrynichus of Bithynia (grammarian 
ii/A.D.) 6, 57 £., 67 £., 72, 76, 81, 91 £., 
Mae. 123, 132, 141, 158, 166, 192, 
195, 198 f.. 221, 272, 279, 315, 351, 
354, 358, 365, 371, 374, 379, 381, 
383, 385, 390, 395 f.. 398, 407, 410 
—see Abbreviations, pp. XxX, Xxiii, 
under Lobeck, Rutherford 

Pindar 277, 304, 326, 365—see also 
Index I (g) p. 512 

Pisidian inscription 177 

Place names 147 

Plato 163, 281, 310, 3438, 378 f., 


(9) Ῥ. 

Plautus 106, 284 

Pleonastic use of ἄρχομαι 455 f. 

Pliny 292 

Plutarch 111, 157, 250, 292, 310, 347, 
362, 365, 370, 379, 386, 390, 392, 
399f., 406, 442, 454—see Index I (9) 
p. 514 

Pollux (Grammaticus ii/A.D.) 346, 365 

Polybius 122 f., 151, 157, 171, 176, 
190, 234, 274, 303, 305, 313, 339, 
343, 347, 379, 384, 390, 392, 396, 
399, 407f., 417f., 457, 464—see 
also Index I (9) p. 514 

Pompeii 80, 94 

Pontic dialect of MGr 43, 58 

Position of vb in sentence 32, 416 

Positive of adj. for superlative 30, 441 

Possessive Compounds (B.V.) 283 

Praeverbia 293 


390, 
5138 


Prepositions 292-328 — Improper 
Prepositions 328 ff.—elision 62— 


Semitic use of 460 ff. 
Present tense 193-207—Aramaising 
use of historic present 456 


SUBJECTS. 541 


Proclitics 54 

Proclus (v/A.D.) 314 

Pronouns: Personal, flexions 179 f.— 
redundant use 43l—unusual fre- 
quency of oblique cases 431—con- 
fusion of personal and demon- 
strative 432—Reflexive 180 f., 432 
—TIndefinite 179—substitutes for 
432—negative of 433—Relative 179 
—in Hebrew and Aramaic con- 
structions 434 ff.—substitutes for 
436—Distributive 437 

Pronunciation 41 ff.—of the vowels 
93 fi.—and Textual Criticism 93 

Proparoxytone 53 f., 328, 337, 371 

Properispomenon 53 f. 

Prothesis 89, 363 

Psalms: Greek of 17f.—see Index 
I (b) p. 501 

Psephism 86 

Pseudo-Anacreon 388 

Psilosis 46, 181, 273, 329 

Punctiliar force in root of πέμπω 253 

Punctuation 46 ff. 


“Q” 7, 11, 20, 28, 30, 183, 203, 313, 
419, 442, 447, 472 

Quality of vowels: change in 94 

Quantity-levelling of vowels : its date 
56, 93 

Quotations, Semitic 153 


Recessive accent 55, 59, 368 

Redundant verb: participle 453— 
auxil. vb 455 

Reduplication 100, 192—double, or 
“ Attic ᾿ 193—irregular redupl. of 
epd. vbs 288—of words to express 
elative force 270, 442 

Reflexive pronouns 180 f. 

Revelation, Book of 3, 16, 33, 480, 484 

Rhaga : transliteration of in Tobit 147 

Rhodes inscription 127 

Root determinatives 351, 364 

Root nouns 332 


Sandhi (i.e. changes in initial and final 
sounds of words when used in a 
word-group or sentence) 130, 300, 
304, 323 

Semantics, 
305, 316 

Semitic colouring of NT Greek 12— 
419—see Semitisms 

Semitic names: declension of 143 ff. 

Semitic quotations 153 

Semitic words : accent of 59 
of 102, 108—transliteration 
106 ff., 110, 143 ff. 


semasiology 267f., 274, 


34, 


spelling 


of 


542 


Semitisms 12-34: see 412—Semit- 
ism defined 14—classified 14 ff., 
477—‘ nests”? of Semitisms in 


Luke 483—Sem. of vocabulary 432, 
481—‘‘ Secondary Semitisms”’ 15, 
26, 414, 432, 437, 440, 451, 460, 470, 
477, 481 

Septuagint 2, 8, 15, 17, 21 ff, 25f., 
28, 30, 34, 65, 67 ff, 71 ff., 83, 90, 
99, 104, 118, 121 ff., 126, 129, 132, 


143-150, 162, 177, 203, 212; 216, 
248, 315, 322, 339, 362, 365f., 384, 
386, 388, 390, 392 ff., 399, 405, 415, 
435, 438, 444, 447, 452 f., 458, 463, 
465 fi., 469, 473, 482 also 
Index 1 (δ) p. 501 

Sextus Empiricus 124, 396—see also 
Index I (g) p. 515 

Shakespeare, 8, 468 

Sibilants, 45, 106 

Silco inscription 202, ae 

Sinaiticus, Codex (δ) 47, 69f., 75 f., 
78, 103, 110, 1138, 129, 170, 191 ἢ 
194, 198, 216, 256, 262, 266, 321, 


395, 434, 437, 465 f. 

pope 276, 378—see also Index 

1 (9) p. 512 

Sophron (Comicus v/B.c.) 105 

Sounds and writing 37-114—classifi- 
cation of sounds 40 

Spelling no longer phonetic 51 

Strabo 79, 85, 151, 278, 341—see also 
Index I (g) p. 514 

Strong Aorist stem 208 


Subjunective:  itacistic confusions 
with indic. 74—forms in contract 
vbs 196, 200 


Suffixes 332—primary and secondary 
333 f. 

Suidas (Lexicographus 
128, 275, 305 

Superlative 164 ff., 3 
stitutes for 442 

Suppletives 167, 235, 238, 252, 255, 
258, 262 

Syllabic augment 189 f., 193 

Syllabification 60 

Symmachus: Greek version of OT 
366 

Syncope 64, 92, 398 

Syracusan Greek 58 

Syriac 433, 436, 445, 445, 449, 453, 
457, 459, 461 f., 464, 467, 470, 473 

Syros: inscription from 86 


- 


x/ACD:)) “72, 


72—Semitic sub- 


Talmud 451, 457 

Targums 450 f., 453, 457, 460, 469 

Tatpurusa (T.P.) Compounds 271, 
367, 370, 390, 392 

Tautology 419 

Tense stems of vbs 185 ff. 

Tertullian 74, 175 


INDEX OF SUBJECTS. 


Testaments of Twelve Patriarchs 362 
—see Index I (c) p. 502 

Testimonia 419, 480 

Textus Receptus (T.R.) 193, 264, 303, 
395, 399 

Thematic verbs 183 ff.—thematising 
tendency in Hellenistic 183, 194, 
212, 245, 251 

Theocritus 122, 273, 292—see Index 
I (g) p. 514 

Theodotion : Greek version of OT (Θ) 
174, 366, 435, 452, 457, 465 

Theognis (Elegiacus vi/B.c.) 428 

Theophanes (chronicler viii/ix A.D.) 
4171. 

Theophrastus (iv/ili B.c.) 124. 153 

Theophylact (xi/xii A.D.) 282 

Thera 86, 124 

Thessalian dialect 108, 304 

Thomas Magister (Grammaticus xili/ 
xiv A.D.) 91. 113, 122, 132, 189, 
326 

Thucydides 378, 387, 417 f.—see also 
Index I (g) p. 513 

Tobit 17—see also Index I (ce) p. 502 

Towns in -a 147 

Transitive force in composition of ἀπό 
298, of κατά 316 f. 

Transliteration οἵ 
106 ff., 110, 143 ff. 


Semitic words 


Uncontracted vowels 90, 120 f., 139 
Unthematic formations receding 183, 
202 


Vaticanus, Codex (B) 15, 47, 56. 69, 
71, 76, 78, 86, 103; 110, 113, 128, 
130, 170, 197, 212, 216, 245, 262, 
395, 434, 437, 458 

Verbals in -réos 188, 224—in -τός 188, 
224, 370, 408 

Verbs : changes in Hellenistic 182 f.— 
Conjugation classes 186 f.—list of 
verbs with prince. pts. 224-266 
—verbal suffixes 381-410—verbs 
coined from prepositional phrases 
408 

Vettius Valens (astrologer) 408—see 
also Index I (g) p. 515 

Vocative : recession of accent 54, 59 
—flexion 118 ff., 129, 134 ff., 137 

Vowels: 42 fii—vowel system in 
Κοινή 64 f.—vowel-gradation 64 f., 
128, 136, 138, 186, 220, 294, 324, 
330, 369, 372—vowel suffixes 334— 
vowel assimilation 378—vowel- 
levelling 302 : see Itacism 


W-sound in Greek 110 

Washington Codex 47f., 61 f., 6b £, 
83, 86 f., 97 f., 100, 103 f., 108, 115; 
419, 181, 212, 234 


INDEX OF SUBJECTS. 


Weak Aorists encroaching on Strong 
Aorist flexions 183—stem 214 

Wesley: John 2, 9, 478—Charles 
154 

Westcott-Hort 38, 48, 50, 58, 61, 69, 
73 1., 76, 78, 81, 87, 98 ff., 103 ff., 
112, 190. 1591 1.15.1Ὁ 146: 149, 
170, 177, 158: Ὁ 101. 107. 1991, 
205, 209, 211, 238, 256, 259, 261, 
265, 269, 273, 294, 315, 339, 348, 
357, 389, 476 

Western Text 31, 69, 107, 133, 211, 
315, 362, 475 


VOL. IJ. PART III.—37 


543 


Word-division 60 
Word-formation 267-410 


Xenophon 151, 158, 290, 339, 343, 
354, 378, 384, 389 ff., 395 f., 399 ff., 
407 f., 417 f., 428, 454 


Zaconian (Laconian) dialect of MGr451 

Zeugma 450 

Zonaras (Lexicographus 1. 
1, A.D:) 57 

Zwaan, J. de: classification of Semit- 
isms 477 


B.C./ 


PRINTED BY 
MORRISON AND GIBB LTD. 
EDINBURGH AND LONDON 


Ψ 


Moulton, James Hope 
A grammar of New Teste- 
ment Greek. dd ed. 


PLEASE DO NOT REMOVE 
CARDS OR SLIPS FROM THIS POCKET 


UNIVERSITY OF TORONTO LIBRARY 


a 


A time ὲ 
wen” eet « 


Ως τὸν τ- 
OD; Sties teres 


Paes WE ra Se PL sere T IE Had Ky ieee pr aNSE SNS HEA τὸ aye ise aniges carga Rate gee ONE 
on ᾿ Be ates SSS “ = 45 ΟΝ antes “ay & 


ΤΥ» ΕΝ ~ 
re ae vs 


ἐν 
ΕΣ 


4 
Se ed — 
SP 


Se 
ω 


“ La abe «Tay 
* 4 me es a 
See Larder ee ΟΝ Ὁ ΣΡ Sa 
τ = ee: : Sore oe ay 
κῃ rine. =. - 2 oe — 3g ο. 4 
PO eel OF So oe! es Shes Shee A es 
- - = : ; x xe "7 Same 
tote feeig he os Aaa 
ox - ΩΣ » 5 - 
Ey = Ὁ: et Sy aos . a 


Ses 
Pa pacomy three τρις ae ws 
PEST ee cata
Internet Archive Audio

Featured

Top

Images

Featured

Top

Software

Featured

Top

Books

Featured

Top

Video

Featured

Top

Mobile Apps

Browser Extensions

Archive-It Subscription

Save Page Now

Full text of "A grammar of New Testament Greek"

See other formats
